{
    "metadata": {
        "benchmark": "Terminal-Bench-Science",
        "slug": "terminal-bench-science",
        "description": "Research workflow tasks contributed by practicing scientists across five domains",
        "benchmark_id": "terminal_bench_science",
        "family": "terminal_bench_science",
        "version": "0.1",
        "updated": "2026-10-25",
        "dataset_type": "public",
        "industry": "science",
        "tasks": {
            "overall": "Overall",
            "life_sciences": "Life Sciences",
            "physical_sciences": "Physical Sciences",
            "earth_sciences": "Earth Sciences",
            "mathematical_sciences": "Mathematical Sciences",
            "engineering_sciences": "Engineering Sciences"
        },
        "models": [
            "alibaba/qwen3.8-27b",
            "alibaba/qwen3.8-max",
            "anthropic/claude-fable-5",
            "anthropic/claude-fable-5-1",
            "anthropic/claude-haiku-5-5",
            "anthropic/claude-opus-4-8",
            "anthropic/claude-opus-5",
            "anthropic/claude-opus-5-5",
            "anthropic/claude-sonnet-5",
            "anthropic/claude-sonnet-5-5",
            "deepseek/deepseek-v4-flash-0731",
            "deepseek/deepseek-v4-pro-0813",
            "deepseek/deepseek-v4.1-flash",
            "fireworks/ember-1",
            "google/gemini-3.1-pro-preview",
            "google/gemini-3.5-flash",
            "google/gemini-3.6-flash",
            "google/gemini-3.7-flash",
            "google/gemini-3.8-flash",
            "google/gemini-4-argon",
            "grok/grok-4.6",
            "grok/grok-4.7",
            "inception/mercury-2.5",
            "kimi/kimi-k3",
            "kimi/kimi-k3.1-pro",
            "meta/muse-wonder",
            "meta/muse-wonder-full",
            "meta/muse_spark_1_3",
            "meta/muse_spark_1_3_max",
            "openai/gpt-5.6-luna",
            "openai/gpt-5.6-sol",
            "openai/gpt-5.6-terra",
            "openai/gpt-6-astra",
            "openai/gpt-6-luna",
            "openai/gpt-6-sol",
            "openai/gpt-6.1-sol",
            "reflection/Beam-1.0",
            "stepfun/step-5-preview",
            "tencent/hy4-preview",
            "xiaomi/mimo-v2.6-flash",
            "xiaomi/mimo-v2.6-pro",
            "zai/glm-5.3",
            "zai/glm-5.3-flash"
        ],
        "partners": [],
        "showBadge": false,
        "visible": true,
        "use_cost_per_test": true,
        "token_tasks_measured": 70,
        "harness": "Mini-SWE-agent",
        "runner": "external",
        "mode": "agentic",
        "archived": false,
        "partner": false,
        "total_models": 43
    },
    "tasks": {
        "overall": {
            "openai/gpt-6-astra": {
                "accuracy": 62.857,
                "latency": 6534.868,
                "stderr": 5.817,
                "cost_per_test": 20.802802,
                "token_totals": {
                    "input_tokens": 405026379,
                    "output_tokens": 11095298,
                    "reasoning_tokens": 8544230,
                    "cache_read_tokens": 386366671,
                    "cache_write_tokens": 18635815
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 47.1,
                    "avg_query_seconds": 120.1,
                    "median_query_seconds": 65.1,
                    "avg_ttft_seconds": 98.6,
                    "output_time_split": {
                        "reasoning_pct": 2.9,
                        "content_pct": 18.6,
                        "tool_call_pct": 78.5
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-6.1-sol": {
                "accuracy": 52.857,
                "latency": 5003.987,
                "stderr": 6.009,
                "cost_per_test": 3.436948,
                "token_totals": {
                    "input_tokens": 337815176,
                    "output_tokens": 11533409,
                    "reasoning_tokens": 9185469,
                    "cache_read_tokens": 317832491,
                    "cache_write_tokens": 19972804
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 42.1,
                    "avg_query_seconds": 130.2,
                    "median_query_seconds": 61.4,
                    "avg_ttft_seconds": 106.4,
                    "output_time_split": {
                        "reasoning_pct": 4.0,
                        "content_pct": 11.1,
                        "tool_call_pct": 84.9
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-opus-5-5": {
                "accuracy": 47.143,
                "latency": 8856.116,
                "stderr": 6.009,
                "cost_per_test": 19.121966,
                "token_totals": {
                    "input_tokens": 1829311051,
                    "output_tokens": 17393519,
                    "reasoning_tokens": 11843007,
                    "cache_read_tokens": 1700954981,
                    "cache_write_tokens": 128339444
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 117.3,
                    "avg_query_seconds": 22.9,
                    "median_query_seconds": 14.2,
                    "avg_ttft_seconds": 21.8,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-sonnet-5-5": {
                "accuracy": 47.143,
                "latency": 9980.148,
                "stderr": 6.009,
                "cost_per_test": 19.568918,
                "token_totals": {
                    "input_tokens": 9111131259,
                    "output_tokens": 34342805,
                    "reasoning_tokens": 24263221,
                    "cache_read_tokens": 9063088192,
                    "cache_write_tokens": 48002507
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 285.1,
                    "avg_query_seconds": 16.6,
                    "median_query_seconds": 10.6,
                    "avg_ttft_seconds": 16.4,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "google/gemini-4-argon": {
                "accuracy": 44.286,
                "latency": 6369.095,
                "stderr": 5.98,
                "cost_per_test": 44.904641,
                "token_totals": {
                    "input_tokens": 2230737485,
                    "output_tokens": 21755185,
                    "reasoning_tokens": 10397231,
                    "cache_read_tokens": 1635454933,
                    "cache_write_tokens": 0
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 113.1,
                    "avg_query_seconds": 47.3,
                    "median_query_seconds": 35.2,
                    "avg_ttft_seconds": 14.4,
                    "output_time_split": {
                        "reasoning_pct": 52.8,
                        "content_pct": 46.9,
                        "tool_call_pct": 0.3
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 262144,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Google",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-fable-5-1": {
                "accuracy": 40.0,
                "latency": 9376.226,
                "stderr": 5.898,
                "cost_per_test": 38.010073,
                "token_totals": {
                    "input_tokens": 1408818997,
                    "output_tokens": 15104550,
                    "reasoning_tokens": 9261638,
                    "cache_read_tokens": 1275387088,
                    "cache_write_tokens": 133417273
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 105.8,
                    "avg_query_seconds": 28.9,
                    "median_query_seconds": 17.9,
                    "avg_ttft_seconds": null,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "meta/muse-wonder-full": {
                "accuracy": 37.143,
                "latency": 12919.209,
                "stderr": 5.817,
                "cost_per_test": 22.814086,
                "token_totals": {
                    "input_tokens": 3713083934,
                    "output_tokens": 17062678,
                    "reasoning_tokens": 11556867,
                    "cache_read_tokens": 3578053496,
                    "cache_write_tokens": 0
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 191.6,
                    "avg_query_seconds": 54.4,
                    "median_query_seconds": 20.5,
                    "avg_ttft_seconds": 52.5,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Meta",
                "harness": "Mini-SWE-agent"
            },
            "kimi/kimi-k3.1-pro": {
                "accuracy": 31.429,
                "latency": 3474.879,
                "stderr": 5.589,
                "cost_per_test": 8.577155,
                "token_totals": {
                    "input_tokens": 2310199935,
                    "output_tokens": 10019167,
                    "reasoning_tokens": 5787120,
                    "cache_read_tokens": 2288994816,
                    "cache_write_tokens": 0
                },
                "token_tasks_measured": 70,
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 262144,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Moonshot AI",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-6-sol": {
                "accuracy": 30.0,
                "latency": 4930.428,
                "stderr": 5.517,
                "cost_per_test": 5.821477,
                "token_totals": {
                    "input_tokens": 1106517592,
                    "output_tokens": 9142533,
                    "reasoning_tokens": 5482665,
                    "cache_read_tokens": 1090637986,
                    "cache_write_tokens": 15843344
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 105.6,
                    "avg_query_seconds": 26.1,
                    "median_query_seconds": 15.6,
                    "avg_ttft_seconds": 13.7,
                    "output_time_split": {
                        "reasoning_pct": 12.8,
                        "content_pct": 5.8,
                        "tool_call_pct": 81.3
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-opus-5": {
                "accuracy": 27.143,
                "latency": 7908.453,
                "stderr": 5.354,
                "cost_per_test": 32.543386,
                "token_totals": {
                    "input_tokens": 3251326101,
                    "output_tokens": 18648623,
                    "reasoning_tokens": 11809096,
                    "cache_read_tokens": 3218944868,
                    "cache_write_tokens": 32354267
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 192.6,
                    "avg_query_seconds": 20.2,
                    "median_query_seconds": 9.2,
                    "avg_ttft_seconds": 19.0,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "meta/muse-wonder": {
                "accuracy": 22.857,
                "latency": 5584.198,
                "stderr": 5.055,
                "cost_per_test": 19.609644,
                "token_totals": {
                    "input_tokens": 4434041175,
                    "output_tokens": 17847376,
                    "reasoning_tokens": 12044271,
                    "cache_read_tokens": 4400115036,
                    "cache_write_tokens": 0
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 238.7,
                    "avg_query_seconds": 13.5,
                    "median_query_seconds": 7.1,
                    "avg_ttft_seconds": 12.3,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Meta",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-5.6-sol": {
                "accuracy": 20.0,
                "latency": 6216.027,
                "stderr": 4.815,
                "cost_per_test": 7.322071,
                "token_totals": {
                    "input_tokens": 662422169,
                    "output_tokens": 6024299,
                    "reasoning_tokens": 2789570,
                    "cache_read_tokens": 648354535,
                    "cache_write_tokens": 14038242
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 72.8,
                    "avg_query_seconds": 19.7,
                    "median_query_seconds": 13.1,
                    "avg_ttft_seconds": 10.0,
                    "output_time_split": {
                        "reasoning_pct": 14.9,
                        "content_pct": 26.8,
                        "tool_call_pct": 58.3
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-fable-5": {
                "accuracy": 15.714,
                "latency": 7157.669,
                "stderr": 4.381,
                "cost_per_test": 51.990885,
                "token_totals": {
                    "input_tokens": 2599459585,
                    "output_tokens": 16914583,
                    "reasoning_tokens": 10535367,
                    "cache_read_tokens": 2569404286,
                    "cache_write_tokens": 30031671
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 166.5,
                    "avg_query_seconds": 24.6,
                    "median_query_seconds": 9.6,
                    "avg_ttft_seconds": 21.9,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "alibaba/qwen3.8-max": {
                "accuracy": 12.857,
                "latency": 12243.338,
                "stderr": 4.03,
                "cost_per_test": 16.682189,
                "token_totals": {
                    "input_tokens": 3844140557,
                    "output_tokens": 22910070,
                    "reasoning_tokens": 16758796,
                    "cache_read_tokens": 3804564736,
                    "cache_write_tokens": null
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 185.4,
                    "avg_query_seconds": 49.2,
                    "median_query_seconds": 31.6,
                    "avg_ttft_seconds": 10.9,
                    "output_time_split": {
                        "reasoning_pct": 77.2,
                        "content_pct": 0.3,
                        "tool_call_pct": 22.5
                    }
                },
                "temperature": 0.7,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Alibaba",
                "harness": "Mini-SWE-agent"
            },
            "grok/grok-4.6": {
                "accuracy": 11.429,
                "latency": 4696.051,
                "stderr": 3.83,
                "cost_per_test": 4.815933,
                "token_totals": {
                    "input_tokens": 519819247,
                    "output_tokens": 7158015,
                    "reasoning_tokens": 3485074,
                    "cache_read_tokens": 496980864,
                    "cache_write_tokens": 0
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 65.2,
                    "avg_query_seconds": 24.2,
                    "median_query_seconds": 13.9,
                    "avg_ttft_seconds": 4.8,
                    "output_time_split": {
                        "reasoning_pct": 79.3,
                        "content_pct": 20.7,
                        "tool_call_pct": 0.0
                    }
                },
                "temperature": 0.7,
                "top_p": 0.95,
                "max_output_tokens": null,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "SpaceXAI",
                "harness": "Mini-SWE-agent"
            },
            "meta/muse_spark_1_3_max": {
                "accuracy": 10.0,
                "latency": 5045.791,
                "stderr": 3.612,
                "cost_per_test": 6.094983,
                "token_totals": {
                    "input_tokens": 1998607677,
                    "output_tokens": 19369661,
                    "reasoning_tokens": 12665202,
                    "cache_read_tokens": 1958119889,
                    "cache_write_tokens": 0
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 132.3,
                    "avg_query_seconds": 13.5,
                    "median_query_seconds": 8.9,
                    "avg_ttft_seconds": 6.0,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Meta",
                "harness": "Mini-SWE-agent"
            },
            "grok/grok-4.7": {
                "accuracy": 10.0,
                "latency": 5553.191,
                "stderr": 3.612,
                "cost_per_test": 17.040638,
                "token_totals": {
                    "input_tokens": 1158107075,
                    "output_tokens": 11957556,
                    "reasoning_tokens": 6886200,
                    "cache_read_tokens": 1084056704,
                    "cache_write_tokens": 0
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 93.7,
                    "avg_query_seconds": 26.6,
                    "median_query_seconds": 18.4,
                    "avg_ttft_seconds": 3.7,
                    "output_time_split": {
                        "reasoning_pct": 94.4,
                        "content_pct": 5.6,
                        "tool_call_pct": 0.0
                    }
                },
                "temperature": 0.7,
                "top_p": 0.95,
                "max_output_tokens": null,
                "reasoning": null,
                "reasoning_effort": "xhigh",
                "verbosity": null,
                "compute_effort": null,
                "provider": "SpaceXAI",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-haiku-5-5": {
                "accuracy": 10.0,
                "latency": 5637.472,
                "stderr": 3.612,
                "cost_per_test": 4.523698,
                "token_totals": {
                    "input_tokens": 4964740201,
                    "output_tokens": 26152873,
                    "reasoning_tokens": 19018280,
                    "cache_read_tokens": 4931170604,
                    "cache_write_tokens": 33540337
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 208.4,
                    "avg_query_seconds": 9.0,
                    "median_query_seconds": 4.1,
                    "avg_ttft_seconds": 7.6,
                    "output_time_split": null
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-5.6-terra": {
                "accuracy": 10.0,
                "latency": 7443.23,
                "stderr": 3.612,
                "cost_per_test": 5.19351,
                "token_totals": {
                    "input_tokens": 935527799,
                    "output_tokens": 7338523,
                    "reasoning_tokens": 4151458,
                    "cache_read_tokens": 919964318,
                    "cache_write_tokens": 15421007
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 96.4,
                    "avg_query_seconds": 19.3,
                    "median_query_seconds": 12.4,
                    "avg_ttft_seconds": 9.1,
                    "output_time_split": {
                        "reasoning_pct": 25.6,
                        "content_pct": 8.9,
                        "tool_call_pct": 65.4
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "google/gemini-3.8-flash": {
                "accuracy": 8.571,
                "latency": 3311.435,
                "stderr": 3.37,
                "cost_per_test": 5.643008,
                "token_totals": {
                    "input_tokens": 1352310020,
                    "output_tokens": 11150231,
                    "reasoning_tokens": 5862209,
                    "cache_read_tokens": 1271911976,
                    "cache_write_tokens": 0
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 69,
                    "avg_turns": 146.3,
                    "avg_query_seconds": 5.8,
                    "median_query_seconds": 4.2,
                    "avg_ttft_seconds": 3.3,
                    "output_time_split": {
                        "reasoning_pct": 97.9,
                        "content_pct": 0.9,
                        "tool_call_pct": 1.3
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Google",
                "harness": "Mini-SWE-agent"
            },
            "google/gemini-3.7-flash": {
                "accuracy": 7.143,
                "latency": 4825.144,
                "stderr": 3.1,
                "cost_per_test": 7.190413,
                "token_totals": {
                    "input_tokens": 1286456118,
                    "output_tokens": 10695665,
                    "reasoning_tokens": 3697777,
                    "cache_read_tokens": 1115979803,
                    "cache_write_tokens": 0
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 109.8,
                    "avg_query_seconds": 6.7,
                    "median_query_seconds": 5.1,
                    "avg_ttft_seconds": 4.3,
                    "output_time_split": {
                        "reasoning_pct": 96.5,
                        "content_pct": 2.0,
                        "tool_call_pct": 1.5
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Google",
                "harness": "Mini-SWE-agent"
            },
            "google/gemini-3.5-flash": {
                "accuracy": 5.714,
                "latency": 3654.61,
                "stderr": 2.794,
                "cost_per_test": 3.802646,
                "token_totals": {
                    "input_tokens": 634004951,
                    "output_tokens": 8326111,
                    "reasoning_tokens": 4021440,
                    "cache_read_tokens": 562783101,
                    "cache_write_tokens": 0
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 83.1,
                    "avg_query_seconds": 8.5,
                    "median_query_seconds": 4.8,
                    "avg_ttft_seconds": 3.1,
                    "output_time_split": {
                        "reasoning_pct": 57.6,
                        "content_pct": 41.8,
                        "tool_call_pct": 0.6
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Google",
                "harness": "Mini-SWE-agent"
            },
            "zai/glm-5.3": {
                "accuracy": 5.714,
                "latency": 9898.858,
                "stderr": 2.794,
                "cost_per_test": 15.627955,
                "token_totals": {
                    "input_tokens": 3692684680,
                    "output_tokens": 18182694,
                    "reasoning_tokens": 13144712,
                    "cache_read_tokens": 3645443456,
                    "cache_write_tokens": 0
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 220.6,
                    "avg_query_seconds": 25.9,
                    "median_query_seconds": 12.6,
                    "avg_ttft_seconds": 6.9,
                    "output_time_split": {
                        "reasoning_pct": 99.1,
                        "content_pct": 0.9,
                        "tool_call_pct": 0.0
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Fireworks AI",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-sonnet-5": {
                "accuracy": 5.714,
                "latency": 11157.924,
                "stderr": 2.794,
                "cost_per_test": 30.299862,
                "token_totals": {
                    "input_tokens": 8555375360,
                    "output_tokens": 26718888,
                    "reasoning_tokens": 18294700,
                    "cache_read_tokens": 8493310382,
                    "cache_write_tokens": 62018810
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 333.3,
                    "avg_query_seconds": 16.1,
                    "median_query_seconds": 7.7,
                    "avg_ttft_seconds": 15.2,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "fireworks/ember-1": {
                "accuracy": 5.714,
                "latency": 12196.691,
                "stderr": 2.794,
                "cost_per_test": 30.760832,
                "token_totals": {
                    "input_tokens": 5574642539,
                    "output_tokens": 19999519,
                    "reasoning_tokens": 13579586,
                    "cache_read_tokens": 5507652660,
                    "cache_write_tokens": 0
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 276.7,
                    "avg_query_seconds": 23.4,
                    "median_query_seconds": 13.2,
                    "avg_ttft_seconds": 6.4,
                    "output_time_split": {
                        "reasoning_pct": 71.1,
                        "content_pct": 1.7,
                        "tool_call_pct": 27.2
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Fireworks AI",
                "harness": "Mini-SWE-agent"
            },
            "meta/muse_spark_1_3": {
                "accuracy": 4.286,
                "latency": 4050.414,
                "stderr": 2.438,
                "cost_per_test": 4.407744,
                "token_totals": {
                    "input_tokens": 1454427801,
                    "output_tokens": 14032439,
                    "reasoning_tokens": 8560708,
                    "cache_read_tokens": 1426482303,
                    "cache_write_tokens": 0
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 124.8,
                    "avg_query_seconds": 9.5,
                    "median_query_seconds": 6.3,
                    "avg_ttft_seconds": 5.1,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": "xhigh",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Meta",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-6-luna": {
                "accuracy": 4.286,
                "latency": 4559.47,
                "stderr": 2.438,
                "cost_per_test": 0.240966,
                "token_totals": {
                    "input_tokens": 771110577,
                    "output_tokens": 9956631,
                    "reasoning_tokens": 7497116,
                    "cache_read_tokens": 755113085,
                    "cache_write_tokens": 15962745
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 76.0,
                    "avg_query_seconds": 20.7,
                    "median_query_seconds": 10.9,
                    "avg_ttft_seconds": 7.2,
                    "output_time_split": {
                        "reasoning_pct": 47.6,
                        "content_pct": 5.1,
                        "tool_call_pct": 47.3
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "deepseek/deepseek-v4-flash-0731": {
                "accuracy": 4.286,
                "latency": 5653.642,
                "stderr": 2.438,
                "cost_per_test": 1.095543,
                "token_totals": {
                    "input_tokens": 3628605043,
                    "output_tokens": 17032330,
                    "reasoning_tokens": 12139861,
                    "cache_read_tokens": 3620612352,
                    "cache_write_tokens": 0
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 228.5,
                    "avg_query_seconds": 7.4,
                    "median_query_seconds": 4.9,
                    "avg_ttft_seconds": 2.7,
                    "output_time_split": {
                        "reasoning_pct": 81.3,
                        "content_pct": 0.5,
                        "tool_call_pct": 18.2
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 384000,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "DeepSeek",
                "harness": "Mini-SWE-agent"
            },
            "google/gemini-3.6-flash": {
                "accuracy": 4.286,
                "latency": 6744.655,
                "stderr": 2.438,
                "cost_per_test": 4.390771,
                "token_totals": {
                    "input_tokens": 906982870,
                    "output_tokens": 10616369,
                    "reasoning_tokens": 3143772,
                    "cache_read_tokens": 839068957,
                    "cache_write_tokens": 0
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 69,
                    "avg_turns": 98.1,
                    "avg_query_seconds": 9.2,
                    "median_query_seconds": 6.2,
                    "avg_ttft_seconds": 3.2,
                    "output_time_split": {
                        "reasoning_pct": 90.0,
                        "content_pct": 9.4,
                        "tool_call_pct": 0.6
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Google",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-opus-4-8": {
                "accuracy": 4.286,
                "latency": 6906.538,
                "stderr": 2.438,
                "cost_per_test": 23.141733,
                "token_totals": {
                    "input_tokens": 1809454503,
                    "output_tokens": 20000133,
                    "reasoning_tokens": 14489581,
                    "cache_read_tokens": 1772026517,
                    "cache_write_tokens": 37411842
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 115.3,
                    "avg_query_seconds": 35.7,
                    "median_query_seconds": 20.0,
                    "avg_ttft_seconds": 29.4,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "deepseek/deepseek-v4-pro-0813": {
                "accuracy": 4.286,
                "latency": 8181.756,
                "stderr": 2.438,
                "cost_per_test": 4.630215,
                "token_totals": {
                    "input_tokens": 5726611218,
                    "output_tokens": 14168265,
                    "reasoning_tokens": 6714937,
                    "cache_read_tokens": 5714042368,
                    "cache_write_tokens": 0
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 353.2,
                    "avg_query_seconds": 9.0,
                    "median_query_seconds": 5.9,
                    "avg_ttft_seconds": 3.0,
                    "output_time_split": {
                        "reasoning_pct": 67.1,
                        "content_pct": 0.5,
                        "tool_call_pct": 32.4
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 384000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "DeepSeek",
                "harness": "Mini-SWE-agent"
            },
            "xiaomi/mimo-v2.6-flash": {
                "accuracy": 4.286,
                "latency": 14326.966,
                "stderr": 2.438,
                "cost_per_test": 0.244346,
                "token_totals": {
                    "input_tokens": 2120700869,
                    "output_tokens": 19847075,
                    "reasoning_tokens": 14752104,
                    "cache_read_tokens": 2079818560,
                    "cache_write_tokens": 0
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 125.8,
                    "avg_query_seconds": 52.4,
                    "median_query_seconds": 29.4,
                    "avg_ttft_seconds": 13.4,
                    "output_time_split": {
                        "reasoning_pct": 85.9,
                        "content_pct": 0.8,
                        "tool_call_pct": 13.3
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Xiaomi",
                "harness": "Mini-SWE-agent"
            },
            "stepfun/step-5-preview": {
                "accuracy": 2.857,
                "latency": 8997.76,
                "stderr": 2.006,
                "cost_per_test": 4.250933,
                "token_totals": {
                    "input_tokens": 4101639238,
                    "output_tokens": 21535694,
                    "reasoning_tokens": 0,
                    "cache_read_tokens": 4065495040,
                    "cache_write_tokens": 0
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 217.8,
                    "avg_query_seconds": 23.0,
                    "median_query_seconds": 11.4,
                    "avg_ttft_seconds": 5.2,
                    "output_time_split": {
                        "reasoning_pct": 80.3,
                        "content_pct": 0.4,
                        "tool_call_pct": 19.3
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 1024000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Stepfun",
                "harness": "Mini-SWE-agent"
            },
            "xiaomi/mimo-v2.6-pro": {
                "accuracy": 2.857,
                "latency": 11504.04,
                "stderr": 2.006,
                "cost_per_test": 0.62088,
                "token_totals": {
                    "input_tokens": 2819218214,
                    "output_tokens": 17831364,
                    "reasoning_tokens": 13255805,
                    "cache_read_tokens": 2777959296,
                    "cache_write_tokens": 0
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 165.9,
                    "avg_query_seconds": 50.7,
                    "median_query_seconds": 26.0,
                    "avg_ttft_seconds": 14.4,
                    "output_time_split": {
                        "reasoning_pct": 85.1,
                        "content_pct": 1.4,
                        "tool_call_pct": 13.5
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Xiaomi",
                "harness": "Mini-SWE-agent"
            },
            "zai/glm-5.3-flash": {
                "accuracy": 2.857,
                "latency": 12461.546,
                "stderr": 2.006,
                "cost_per_test": 0.524048,
                "token_totals": {
                    "input_tokens": 1931171934,
                    "output_tokens": 17549106,
                    "reasoning_tokens": 13130932,
                    "cache_read_tokens": 1875697280,
                    "cache_write_tokens": 0
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 141.9,
                    "avg_query_seconds": 39.6,
                    "median_query_seconds": 21.5,
                    "avg_ttft_seconds": 6.3,
                    "output_time_split": {
                        "reasoning_pct": 98.9,
                        "content_pct": 1.1,
                        "tool_call_pct": 0.0
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Zhipu AI",
                "harness": "Mini-SWE-agent"
            },
            "google/gemini-3.1-pro-preview": {
                "accuracy": 1.429,
                "latency": 2638.065,
                "stderr": 1.429,
                "cost_per_test": 5.907095,
                "token_totals": {
                    "input_tokens": 858729252,
                    "output_tokens": 3958785,
                    "reasoning_tokens": 2054169,
                    "cache_read_tokens": 796983036,
                    "cache_write_tokens": 0
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 122.2,
                    "avg_query_seconds": 12.0,
                    "median_query_seconds": 7.1,
                    "avg_ttft_seconds": 9.7,
                    "output_time_split": {
                        "reasoning_pct": 84.7,
                        "content_pct": 14.3,
                        "tool_call_pct": 1.0
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Google",
                "harness": "Mini-SWE-agent"
            },
            "reflection/Beam-1.0": {
                "accuracy": 1.429,
                "latency": 4658.837,
                "stderr": 1.429,
                "cost_per_test": null,
                "token_totals": {
                    "input_tokens": 1817962438,
                    "output_tokens": 4532859,
                    "reasoning_tokens": 2389890,
                    "cache_read_tokens": 1800613696,
                    "cache_write_tokens": 0
                },
                "token_tasks_measured": 70,
                "temperature": 0.7,
                "top_p": 0.99,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Reflection",
                "harness": "Mini-SWE-agent"
            },
            "tencent/hy4-preview": {
                "accuracy": 1.429,
                "latency": 13679.187,
                "stderr": 1.429,
                "cost_per_test": 2.143682,
                "token_totals": {
                    "input_tokens": 1954683816,
                    "output_tokens": 18902210,
                    "reasoning_tokens": 11136018,
                    "cache_read_tokens": 1928564352,
                    "cache_write_tokens": 0
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 174.0,
                    "avg_query_seconds": 31.1,
                    "median_query_seconds": 10.7,
                    "avg_ttft_seconds": 5.5,
                    "output_time_split": {
                        "reasoning_pct": 64.8,
                        "content_pct": 35.2,
                        "tool_call_pct": 0.0
                    }
                },
                "temperature": 0.9,
                "top_p": 1.0,
                "max_output_tokens": 64000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Tencent",
                "harness": "Terminus 2"
            },
            "kimi/kimi-k3": {
                "accuracy": 1.429,
                "latency": 17349.404,
                "stderr": 1.429,
                "cost_per_test": 21.301409,
                "token_totals": {
                    "input_tokens": 3528346373,
                    "output_tokens": 19628672,
                    "reasoning_tokens": 13477143,
                    "cache_read_tokens": 3477174272,
                    "cache_write_tokens": 49305480
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 213.9,
                    "avg_query_seconds": 57.6,
                    "median_query_seconds": 35.7,
                    "avg_ttft_seconds": 22.3,
                    "output_time_split": {
                        "reasoning_pct": 74.5,
                        "content_pct": 2.4,
                        "tool_call_pct": 23.1
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 262144,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Moonshot AI",
                "harness": "Mini-SWE-agent"
            },
            "inception/mercury-2.5": {
                "accuracy": 0.0,
                "latency": 1693.946,
                "stderr": 0.0,
                "cost_per_test": 0.17804,
                "token_totals": {
                    "input_tokens": 256765973,
                    "output_tokens": 3081607,
                    "reasoning_tokens": 2995809,
                    "cache_read_tokens": 228897807,
                    "cache_write_tokens": 0
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 76.4,
                    "avg_query_seconds": 2.5,
                    "median_query_seconds": 1.8,
                    "avg_ttft_seconds": 1.6,
                    "output_time_split": null
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Inception",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-5.6-luna": {
                "accuracy": 0.0,
                "latency": 7396.452,
                "stderr": 0.0,
                "cost_per_test": 0.581456,
                "token_totals": {
                    "input_tokens": 1129716462,
                    "output_tokens": 7266560,
                    "reasoning_tokens": 3921207,
                    "cache_read_tokens": 1113668160,
                    "cache_write_tokens": 16009244
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 120.9,
                    "avg_query_seconds": 10.7,
                    "median_query_seconds": 6.9,
                    "avg_ttft_seconds": 5.7,
                    "output_time_split": {
                        "reasoning_pct": 22.5,
                        "content_pct": 1.4,
                        "tool_call_pct": 76.1
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "deepseek/deepseek-v4.1-flash": {
                "accuracy": 0.0,
                "latency": 10621.573,
                "stderr": 0.0,
                "cost_per_test": 0.473154,
                "token_totals": {
                    "input_tokens": 1765162289,
                    "output_tokens": 17312542,
                    "reasoning_tokens": 12785452,
                    "cache_read_tokens": 1759193848,
                    "cache_write_tokens": 0
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 136.1,
                    "avg_query_seconds": 10.0,
                    "median_query_seconds": 7.1,
                    "avg_ttft_seconds": 2.3,
                    "output_time_split": {
                        "reasoning_pct": 82.0,
                        "content_pct": 0.4,
                        "tool_call_pct": 17.6
                    }
                },
                "temperature": 1.0,
                "top_p": 1.0,
                "max_output_tokens": 384000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "DeepSeek",
                "harness": "Mini-SWE-agent"
            },
            "alibaba/qwen3.8-27b": {
                "accuracy": 0.0,
                "latency": 14549.775,
                "stderr": 0.0,
                "cost_per_test": 11.135637,
                "token_totals": {
                    "input_tokens": 9837173014,
                    "output_tokens": 30937435,
                    "reasoning_tokens": 20965925,
                    "cache_read_tokens": 9404231680,
                    "cache_write_tokens": null
                },
                "token_tasks_measured": 70,
                "turn_metrics": {
                    "tasks_measured": 70,
                    "avg_turns": 334.7,
                    "avg_query_seconds": 29.3,
                    "median_query_seconds": 13.5,
                    "avg_ttft_seconds": 16.7,
                    "output_time_split": {
                        "reasoning_pct": 75.6,
                        "content_pct": 2.0,
                        "tool_call_pct": 22.4
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "xhigh",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Alibaba",
                "harness": "Mini-SWE-agent"
            }
        },
        "life_sciences": {
            "openai/gpt-6-astra": {
                "accuracy": 68.421,
                "latency": null,
                "stderr": 10.956,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 43.2,
                    "avg_query_seconds": 104.5,
                    "median_query_seconds": 43.7,
                    "avg_ttft_seconds": 88.5,
                    "output_time_split": {
                        "reasoning_pct": 3.9,
                        "content_pct": 23.3,
                        "tool_call_pct": 72.8
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-opus-5-5": {
                "accuracy": 52.632,
                "latency": null,
                "stderr": 11.769,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 112.1,
                    "avg_query_seconds": 20.8,
                    "median_query_seconds": 14.6,
                    "avg_ttft_seconds": 20.1,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-6.1-sol": {
                "accuracy": 47.368,
                "latency": null,
                "stderr": 11.769,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 38.3,
                    "avg_query_seconds": 102.7,
                    "median_query_seconds": 46.6,
                    "avg_ttft_seconds": 84.2,
                    "output_time_split": {
                        "reasoning_pct": 5.1,
                        "content_pct": 16.0,
                        "tool_call_pct": 78.9
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-sonnet-5-5": {
                "accuracy": 42.105,
                "latency": null,
                "stderr": 11.637,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 232.4,
                    "avg_query_seconds": 18.1,
                    "median_query_seconds": 12.6,
                    "avg_ttft_seconds": 17.9,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "google/gemini-4-argon": {
                "accuracy": 42.105,
                "latency": null,
                "stderr": 11.637,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 108.9,
                    "avg_query_seconds": 39.3,
                    "median_query_seconds": 33.1,
                    "avg_ttft_seconds": 12.5,
                    "output_time_split": {
                        "reasoning_pct": 55.6,
                        "content_pct": 44.1,
                        "tool_call_pct": 0.3
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 262144,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Google",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-6-sol": {
                "accuracy": 42.105,
                "latency": null,
                "stderr": 11.637,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 76.7,
                    "avg_query_seconds": 31.5,
                    "median_query_seconds": 18.1,
                    "avg_ttft_seconds": 14.7,
                    "output_time_split": {
                        "reasoning_pct": 8.8,
                        "content_pct": 4.1,
                        "tool_call_pct": 87.1
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "kimi/kimi-k3.1-pro": {
                "accuracy": 36.842,
                "latency": null,
                "stderr": 11.37,
                "cost_per_test": null,
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 262144,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Moonshot AI",
                "harness": "Mini-SWE-agent"
            },
            "meta/muse-wonder-full": {
                "accuracy": 36.842,
                "latency": null,
                "stderr": 11.37,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 167.6,
                    "avg_query_seconds": 64.3,
                    "median_query_seconds": 30.2,
                    "avg_ttft_seconds": 62.1,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Meta",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-opus-5": {
                "accuracy": 31.579,
                "latency": null,
                "stderr": 10.956,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 163.9,
                    "avg_query_seconds": 18.5,
                    "median_query_seconds": 10.2,
                    "avg_ttft_seconds": 17.4,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-fable-5-1": {
                "accuracy": 26.316,
                "latency": null,
                "stderr": 10.379,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 97.1,
                    "avg_query_seconds": 27.8,
                    "median_query_seconds": 19.5,
                    "avg_ttft_seconds": 26.5,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "grok/grok-4.6": {
                "accuracy": 21.053,
                "latency": null,
                "stderr": 9.609,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 46.7,
                    "avg_query_seconds": 23.6,
                    "median_query_seconds": 15.2,
                    "avg_ttft_seconds": 4.2,
                    "output_time_split": {
                        "reasoning_pct": 77.5,
                        "content_pct": 22.4,
                        "tool_call_pct": 0.0
                    }
                },
                "temperature": 0.7,
                "top_p": 0.95,
                "max_output_tokens": null,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "SpaceXAI",
                "harness": "Mini-SWE-agent"
            },
            "meta/muse-wonder": {
                "accuracy": 21.053,
                "latency": null,
                "stderr": 9.609,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 204.2,
                    "avg_query_seconds": 13.7,
                    "median_query_seconds": 8.5,
                    "avg_ttft_seconds": 12.5,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Meta",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-5.6-sol": {
                "accuracy": 21.053,
                "latency": null,
                "stderr": 9.609,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 56.6,
                    "avg_query_seconds": 19.9,
                    "median_query_seconds": 14.4,
                    "avg_ttft_seconds": 10.3,
                    "output_time_split": {
                        "reasoning_pct": 15.1,
                        "content_pct": 33.6,
                        "tool_call_pct": 51.3
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-haiku-5-5": {
                "accuracy": 15.789,
                "latency": null,
                "stderr": 8.595,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 189.5,
                    "avg_query_seconds": 8.8,
                    "median_query_seconds": 4.7,
                    "avg_ttft_seconds": 7.3,
                    "output_time_split": null
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "meta/muse_spark_1_3": {
                "accuracy": 15.789,
                "latency": null,
                "stderr": 8.595,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 98.3,
                    "avg_query_seconds": 8.6,
                    "median_query_seconds": 5.8,
                    "avg_ttft_seconds": 4.9,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": "xhigh",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Meta",
                "harness": "Mini-SWE-agent"
            },
            "meta/muse_spark_1_3_max": {
                "accuracy": 15.789,
                "latency": null,
                "stderr": 8.595,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 118.2,
                    "avg_query_seconds": 11.2,
                    "median_query_seconds": 7.7,
                    "avg_ttft_seconds": 5.6,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Meta",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-5.6-terra": {
                "accuracy": 15.789,
                "latency": null,
                "stderr": 8.595,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 75.6,
                    "avg_query_seconds": 19.7,
                    "median_query_seconds": 14.7,
                    "avg_ttft_seconds": 8.9,
                    "output_time_split": {
                        "reasoning_pct": 26.4,
                        "content_pct": 8.7,
                        "tool_call_pct": 64.9
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "zai/glm-5.3": {
                "accuracy": 15.789,
                "latency": null,
                "stderr": 8.595,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 200.4,
                    "avg_query_seconds": 23.6,
                    "median_query_seconds": 13.5,
                    "avg_ttft_seconds": 6.6,
                    "output_time_split": {
                        "reasoning_pct": 99.1,
                        "content_pct": 0.9,
                        "tool_call_pct": 0.0
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Fireworks AI",
                "harness": "Mini-SWE-agent"
            },
            "alibaba/qwen3.8-max": {
                "accuracy": 10.526,
                "latency": null,
                "stderr": 7.234,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 153.2,
                    "avg_query_seconds": 47.8,
                    "median_query_seconds": 33.4,
                    "avg_ttft_seconds": 10.3,
                    "output_time_split": {
                        "reasoning_pct": 76.6,
                        "content_pct": 0.3,
                        "tool_call_pct": 23.0
                    }
                },
                "temperature": 0.7,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Alibaba",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-fable-5": {
                "accuracy": 10.526,
                "latency": null,
                "stderr": 7.234,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 133.2,
                    "avg_query_seconds": 22.4,
                    "median_query_seconds": 9.2,
                    "avg_ttft_seconds": 19.6,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-sonnet-5": {
                "accuracy": 10.526,
                "latency": null,
                "stderr": 7.234,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 247.1,
                    "avg_query_seconds": 16.2,
                    "median_query_seconds": 7.8,
                    "avg_ttft_seconds": 15.2,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "google/gemini-3.6-flash": {
                "accuracy": 10.526,
                "latency": null,
                "stderr": 7.234,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 77.1,
                    "avg_query_seconds": 6.3,
                    "median_query_seconds": 4.9,
                    "avg_ttft_seconds": 2.7,
                    "output_time_split": {
                        "reasoning_pct": 81.5,
                        "content_pct": 17.8,
                        "tool_call_pct": 0.7
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Google",
                "harness": "Mini-SWE-agent"
            },
            "google/gemini-3.7-flash": {
                "accuracy": 10.526,
                "latency": null,
                "stderr": 7.234,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 87.2,
                    "avg_query_seconds": 5.2,
                    "median_query_seconds": 4.2,
                    "avg_ttft_seconds": 3.7,
                    "output_time_split": {
                        "reasoning_pct": 94.7,
                        "content_pct": 3.7,
                        "tool_call_pct": 1.6
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Google",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-6-luna": {
                "accuracy": 10.526,
                "latency": null,
                "stderr": 7.234,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 61.6,
                    "avg_query_seconds": 22.3,
                    "median_query_seconds": 12.7,
                    "avg_ttft_seconds": 7.5,
                    "output_time_split": {
                        "reasoning_pct": 51.2,
                        "content_pct": 7.3,
                        "tool_call_pct": 41.5
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "xiaomi/mimo-v2.6-pro": {
                "accuracy": 10.526,
                "latency": null,
                "stderr": 7.234,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 115.5,
                    "avg_query_seconds": 54.6,
                    "median_query_seconds": 31.8,
                    "avg_ttft_seconds": 16.6,
                    "output_time_split": {
                        "reasoning_pct": 84.9,
                        "content_pct": 1.6,
                        "tool_call_pct": 13.5
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Xiaomi",
                "harness": "Mini-SWE-agent"
            },
            "deepseek/deepseek-v4-pro-0813": {
                "accuracy": 5.263,
                "latency": null,
                "stderr": 5.263,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 270.8,
                    "avg_query_seconds": 8.7,
                    "median_query_seconds": 6.4,
                    "avg_ttft_seconds": 2.7,
                    "output_time_split": {
                        "reasoning_pct": 64.4,
                        "content_pct": 0.7,
                        "tool_call_pct": 34.9
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 384000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "DeepSeek",
                "harness": "Mini-SWE-agent"
            },
            "fireworks/ember-1": {
                "accuracy": 5.263,
                "latency": null,
                "stderr": 5.263,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 120.4,
                    "avg_query_seconds": 29.3,
                    "median_query_seconds": 19.7,
                    "avg_ttft_seconds": 7.2,
                    "output_time_split": {
                        "reasoning_pct": 72.4,
                        "content_pct": 1.5,
                        "tool_call_pct": 26.0
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Fireworks AI",
                "harness": "Mini-SWE-agent"
            },
            "google/gemini-3.5-flash": {
                "accuracy": 5.263,
                "latency": null,
                "stderr": 5.263,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 69.9,
                    "avg_query_seconds": 6.4,
                    "median_query_seconds": 4.0,
                    "avg_ttft_seconds": 2.7,
                    "output_time_split": {
                        "reasoning_pct": 64.5,
                        "content_pct": 34.8,
                        "tool_call_pct": 0.7
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Google",
                "harness": "Mini-SWE-agent"
            },
            "google/gemini-3.8-flash": {
                "accuracy": 5.263,
                "latency": null,
                "stderr": 5.263,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 133.9,
                    "avg_query_seconds": 5.1,
                    "median_query_seconds": 4.0,
                    "avg_ttft_seconds": 2.9,
                    "output_time_split": {
                        "reasoning_pct": 98.7,
                        "content_pct": 0.0,
                        "tool_call_pct": 1.3
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Google",
                "harness": "Mini-SWE-agent"
            },
            "grok/grok-4.7": {
                "accuracy": 5.263,
                "latency": null,
                "stderr": 5.263,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 76.2,
                    "avg_query_seconds": 23.7,
                    "median_query_seconds": 18.6,
                    "avg_ttft_seconds": 2.8,
                    "output_time_split": {
                        "reasoning_pct": 94.2,
                        "content_pct": 5.8,
                        "tool_call_pct": 0.0
                    }
                },
                "temperature": 0.7,
                "top_p": 0.95,
                "max_output_tokens": null,
                "reasoning": null,
                "reasoning_effort": "xhigh",
                "verbosity": null,
                "compute_effort": null,
                "provider": "SpaceXAI",
                "harness": "Mini-SWE-agent"
            },
            "xiaomi/mimo-v2.6-flash": {
                "accuracy": 5.263,
                "latency": null,
                "stderr": 5.263,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 98.0,
                    "avg_query_seconds": 49.4,
                    "median_query_seconds": 33.4,
                    "avg_ttft_seconds": 10.9,
                    "output_time_split": {
                        "reasoning_pct": 83.7,
                        "content_pct": 1.1,
                        "tool_call_pct": 15.2
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Xiaomi",
                "harness": "Mini-SWE-agent"
            },
            "zai/glm-5.3-flash": {
                "accuracy": 5.263,
                "latency": null,
                "stderr": 5.263,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 125.0,
                    "avg_query_seconds": 31.9,
                    "median_query_seconds": 20.1,
                    "avg_ttft_seconds": 5.4,
                    "output_time_split": {
                        "reasoning_pct": 99.4,
                        "content_pct": 0.6,
                        "tool_call_pct": 0.0
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Zhipu AI",
                "harness": "Mini-SWE-agent"
            },
            "alibaba/qwen3.8-27b": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 168.0,
                    "avg_query_seconds": 19.9,
                    "median_query_seconds": 11.0,
                    "avg_ttft_seconds": 10.7,
                    "output_time_split": {
                        "reasoning_pct": 72.6,
                        "content_pct": 2.2,
                        "tool_call_pct": 25.2
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "xhigh",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Alibaba",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-opus-4-8": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 87.2,
                    "avg_query_seconds": 35.5,
                    "median_query_seconds": 22.0,
                    "avg_ttft_seconds": 28.8,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "deepseek/deepseek-v4-flash-0731": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 167.5,
                    "avg_query_seconds": 7.2,
                    "median_query_seconds": 5.1,
                    "avg_ttft_seconds": 2.4,
                    "output_time_split": {
                        "reasoning_pct": 81.4,
                        "content_pct": 0.5,
                        "tool_call_pct": 18.1
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 384000,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "DeepSeek",
                "harness": "Mini-SWE-agent"
            },
            "deepseek/deepseek-v4.1-flash": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 102.5,
                    "avg_query_seconds": 9.5,
                    "median_query_seconds": 7.2,
                    "avg_ttft_seconds": 2.0,
                    "output_time_split": {
                        "reasoning_pct": 82.0,
                        "content_pct": 0.6,
                        "tool_call_pct": 17.4
                    }
                },
                "temperature": 1.0,
                "top_p": 1.0,
                "max_output_tokens": 384000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "DeepSeek",
                "harness": "Mini-SWE-agent"
            },
            "google/gemini-3.1-pro-preview": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 74.5,
                    "avg_query_seconds": 13.7,
                    "median_query_seconds": 6.5,
                    "avg_ttft_seconds": 11.1,
                    "output_time_split": {
                        "reasoning_pct": 77.2,
                        "content_pct": 22.1,
                        "tool_call_pct": 0.6
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Google",
                "harness": "Mini-SWE-agent"
            },
            "inception/mercury-2.5": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 57.4,
                    "avg_query_seconds": 2.2,
                    "median_query_seconds": 1.5,
                    "avg_ttft_seconds": 1.4,
                    "output_time_split": null
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Inception",
                "harness": "Mini-SWE-agent"
            },
            "kimi/kimi-k3": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 139.3,
                    "avg_query_seconds": 46.2,
                    "median_query_seconds": 30.2,
                    "avg_ttft_seconds": 15.3,
                    "output_time_split": {
                        "reasoning_pct": 73.9,
                        "content_pct": 2.5,
                        "tool_call_pct": 23.6
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 262144,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Moonshot AI",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-5.6-luna": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 106.1,
                    "avg_query_seconds": 11.4,
                    "median_query_seconds": 7.0,
                    "avg_ttft_seconds": 5.1,
                    "output_time_split": {
                        "reasoning_pct": 18.1,
                        "content_pct": 1.2,
                        "tool_call_pct": 80.7
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "reflection/Beam-1.0": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "temperature": 0.7,
                "top_p": 0.99,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Reflection",
                "harness": "Mini-SWE-agent"
            },
            "stepfun/step-5-preview": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 175.2,
                    "avg_query_seconds": 22.9,
                    "median_query_seconds": 12.8,
                    "avg_ttft_seconds": 4.5,
                    "output_time_split": {
                        "reasoning_pct": 79.5,
                        "content_pct": 0.5,
                        "tool_call_pct": 20.0
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 1024000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Stepfun",
                "harness": "Mini-SWE-agent"
            },
            "tencent/hy4-preview": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 19,
                    "avg_turns": 114.1,
                    "avg_query_seconds": 25.8,
                    "median_query_seconds": 12.2,
                    "avg_ttft_seconds": 4.4,
                    "output_time_split": {
                        "reasoning_pct": 56.0,
                        "content_pct": 44.0,
                        "tool_call_pct": 0.0
                    }
                },
                "temperature": 0.9,
                "top_p": 1.0,
                "max_output_tokens": 64000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Tencent",
                "harness": "Terminus 2"
            }
        },
        "physical_sciences": {
            "anthropic/claude-sonnet-5-5": {
                "accuracy": 58.824,
                "latency": null,
                "stderr": 12.304,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 290.2,
                    "avg_query_seconds": 16.6,
                    "median_query_seconds": 10.2,
                    "avg_ttft_seconds": 16.4,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-opus-5-5": {
                "accuracy": 47.059,
                "latency": null,
                "stderr": 12.478,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 123.4,
                    "avg_query_seconds": 24.2,
                    "median_query_seconds": 14.4,
                    "avg_ttft_seconds": 23.0,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "kimi/kimi-k3.1-pro": {
                "accuracy": 47.059,
                "latency": null,
                "stderr": 12.478,
                "cost_per_test": null,
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 262144,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Moonshot AI",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-6-astra": {
                "accuracy": 47.059,
                "latency": null,
                "stderr": 12.478,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 46.8,
                    "avg_query_seconds": 153.3,
                    "median_query_seconds": 80.2,
                    "avg_ttft_seconds": 130.2,
                    "output_time_split": {
                        "reasoning_pct": 2.5,
                        "content_pct": 7.3,
                        "tool_call_pct": 90.2
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "google/gemini-4-argon": {
                "accuracy": 41.176,
                "latency": null,
                "stderr": 12.304,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 105.9,
                    "avg_query_seconds": 45.0,
                    "median_query_seconds": 32.8,
                    "avg_ttft_seconds": 13.5,
                    "output_time_split": {
                        "reasoning_pct": 53.0,
                        "content_pct": 46.7,
                        "tool_call_pct": 0.3
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 262144,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Google",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-6.1-sol": {
                "accuracy": 41.176,
                "latency": null,
                "stderr": 12.304,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 39.3,
                    "avg_query_seconds": 141.7,
                    "median_query_seconds": 67.6,
                    "avg_ttft_seconds": 116.6,
                    "output_time_split": {
                        "reasoning_pct": 3.2,
                        "content_pct": 9.9,
                        "tool_call_pct": 86.9
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-fable-5-1": {
                "accuracy": 35.294,
                "latency": null,
                "stderr": 11.947,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 106.0,
                    "avg_query_seconds": 29.1,
                    "median_query_seconds": 17.4,
                    "avg_ttft_seconds": null,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-opus-5": {
                "accuracy": 35.294,
                "latency": null,
                "stderr": 11.947,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 217.3,
                    "avg_query_seconds": 20.1,
                    "median_query_seconds": 8.1,
                    "avg_ttft_seconds": 18.7,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "meta/muse-wonder-full": {
                "accuracy": 35.294,
                "latency": null,
                "stderr": 11.947,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 191.9,
                    "avg_query_seconds": 48.9,
                    "median_query_seconds": 16.3,
                    "avg_ttft_seconds": 47.2,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Meta",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-fable-5": {
                "accuracy": 29.412,
                "latency": null,
                "stderr": 11.391,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 189.6,
                    "avg_query_seconds": 24.0,
                    "median_query_seconds": 9.0,
                    "avg_ttft_seconds": 21.4,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-5.6-sol": {
                "accuracy": 29.412,
                "latency": null,
                "stderr": 11.391,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 61.6,
                    "avg_query_seconds": 20.3,
                    "median_query_seconds": 13.1,
                    "avg_ttft_seconds": 9.6,
                    "output_time_split": {
                        "reasoning_pct": 12.4,
                        "content_pct": 17.9,
                        "tool_call_pct": 69.7
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "alibaba/qwen3.8-max": {
                "accuracy": 23.529,
                "latency": null,
                "stderr": 10.605,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 213.2,
                    "avg_query_seconds": 50.3,
                    "median_query_seconds": 31.0,
                    "avg_ttft_seconds": 10.8,
                    "output_time_split": {
                        "reasoning_pct": 78.9,
                        "content_pct": 0.4,
                        "tool_call_pct": 20.8
                    }
                },
                "temperature": 0.7,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Alibaba",
                "harness": "Mini-SWE-agent"
            },
            "grok/grok-4.7": {
                "accuracy": 23.529,
                "latency": null,
                "stderr": 10.605,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 93.9,
                    "avg_query_seconds": 29.4,
                    "median_query_seconds": 18.8,
                    "avg_ttft_seconds": 4.4,
                    "output_time_split": {
                        "reasoning_pct": 94.1,
                        "content_pct": 5.9,
                        "tool_call_pct": 0.0
                    }
                },
                "temperature": 0.7,
                "top_p": 0.95,
                "max_output_tokens": null,
                "reasoning": null,
                "reasoning_effort": "xhigh",
                "verbosity": null,
                "compute_effort": null,
                "provider": "SpaceXAI",
                "harness": "Mini-SWE-agent"
            },
            "meta/muse-wonder": {
                "accuracy": 23.529,
                "latency": null,
                "stderr": 10.605,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 302.2,
                    "avg_query_seconds": 10.9,
                    "median_query_seconds": 5.6,
                    "avg_ttft_seconds": 9.8,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Meta",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-6-sol": {
                "accuracy": 23.529,
                "latency": null,
                "stderr": 10.605,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 93.6,
                    "avg_query_seconds": 25.1,
                    "median_query_seconds": 15.4,
                    "avg_ttft_seconds": 13.8,
                    "output_time_split": {
                        "reasoning_pct": 13.9,
                        "content_pct": 7.0,
                        "tool_call_pct": 79.1
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "google/gemini-3.7-flash": {
                "accuracy": 17.647,
                "latency": null,
                "stderr": 9.531,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 120.6,
                    "avg_query_seconds": 6.5,
                    "median_query_seconds": 5.0,
                    "avg_ttft_seconds": 4.5,
                    "output_time_split": {
                        "reasoning_pct": 98.4,
                        "content_pct": 0.0,
                        "tool_call_pct": 1.6
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Google",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-haiku-5-5": {
                "accuracy": 11.765,
                "latency": null,
                "stderr": 8.055,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 256.4,
                    "avg_query_seconds": 8.1,
                    "median_query_seconds": 3.8,
                    "avg_ttft_seconds": 6.8,
                    "output_time_split": null
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-opus-4-8": {
                "accuracy": 11.765,
                "latency": null,
                "stderr": 8.055,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 175.6,
                    "avg_query_seconds": 29.2,
                    "median_query_seconds": 13.9,
                    "avg_ttft_seconds": 24.4,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-sonnet-5": {
                "accuracy": 11.765,
                "latency": null,
                "stderr": 8.055,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 480.4,
                    "avg_query_seconds": 13.1,
                    "median_query_seconds": 7.1,
                    "avg_ttft_seconds": 12.4,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "deepseek/deepseek-v4-flash-0731": {
                "accuracy": 11.765,
                "latency": null,
                "stderr": 8.055,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 327.8,
                    "avg_query_seconds": 6.6,
                    "median_query_seconds": 4.5,
                    "avg_ttft_seconds": 2.9,
                    "output_time_split": {
                        "reasoning_pct": 81.9,
                        "content_pct": 0.4,
                        "tool_call_pct": 17.7
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 384000,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "DeepSeek",
                "harness": "Mini-SWE-agent"
            },
            "google/gemini-3.8-flash": {
                "accuracy": 11.765,
                "latency": null,
                "stderr": 8.055,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 160.7,
                    "avg_query_seconds": 5.4,
                    "median_query_seconds": 4.1,
                    "avg_ttft_seconds": 3.1,
                    "output_time_split": {
                        "reasoning_pct": 98.6,
                        "content_pct": 0.0,
                        "tool_call_pct": 1.4
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Google",
                "harness": "Mini-SWE-agent"
            },
            "grok/grok-4.6": {
                "accuracy": 11.765,
                "latency": null,
                "stderr": 8.055,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 64.9,
                    "avg_query_seconds": 24.9,
                    "median_query_seconds": 13.3,
                    "avg_ttft_seconds": 4.6,
                    "output_time_split": {
                        "reasoning_pct": 77.1,
                        "content_pct": 22.8,
                        "tool_call_pct": 0.0
                    }
                },
                "temperature": 0.7,
                "top_p": 0.95,
                "max_output_tokens": null,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "SpaceXAI",
                "harness": "Mini-SWE-agent"
            },
            "meta/muse_spark_1_3_max": {
                "accuracy": 11.765,
                "latency": null,
                "stderr": 8.055,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 144.1,
                    "avg_query_seconds": 13.9,
                    "median_query_seconds": 9.8,
                    "avg_ttft_seconds": 6.1,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Meta",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-5.6-terra": {
                "accuracy": 11.765,
                "latency": null,
                "stderr": 8.055,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 89.6,
                    "avg_query_seconds": 19.1,
                    "median_query_seconds": 12.8,
                    "avg_ttft_seconds": 8.9,
                    "output_time_split": {
                        "reasoning_pct": 23.0,
                        "content_pct": 8.8,
                        "tool_call_pct": 68.1
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "stepfun/step-5-preview": {
                "accuracy": 11.765,
                "latency": null,
                "stderr": 8.055,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 268.6,
                    "avg_query_seconds": 21.1,
                    "median_query_seconds": 9.7,
                    "avg_ttft_seconds": 5.3,
                    "output_time_split": {
                        "reasoning_pct": 80.4,
                        "content_pct": 0.3,
                        "tool_call_pct": 19.4
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 1024000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Stepfun",
                "harness": "Mini-SWE-agent"
            },
            "deepseek/deepseek-v4-pro-0813": {
                "accuracy": 5.882,
                "latency": null,
                "stderr": 5.882,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 499.7,
                    "avg_query_seconds": 8.1,
                    "median_query_seconds": 5.3,
                    "avg_ttft_seconds": 3.2,
                    "output_time_split": {
                        "reasoning_pct": 68.1,
                        "content_pct": 0.4,
                        "tool_call_pct": 31.5
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 384000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "DeepSeek",
                "harness": "Mini-SWE-agent"
            },
            "fireworks/ember-1": {
                "accuracy": 5.882,
                "latency": null,
                "stderr": 5.882,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 403.5,
                    "avg_query_seconds": 18.6,
                    "median_query_seconds": 10.0,
                    "avg_ttft_seconds": 5.9,
                    "output_time_split": {
                        "reasoning_pct": 69.7,
                        "content_pct": 1.9,
                        "tool_call_pct": 28.4
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Fireworks AI",
                "harness": "Mini-SWE-agent"
            },
            "google/gemini-3.1-pro-preview": {
                "accuracy": 5.882,
                "latency": null,
                "stderr": 5.882,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 228.2,
                    "avg_query_seconds": 11.1,
                    "median_query_seconds": 6.9,
                    "avg_ttft_seconds": 9.5,
                    "output_time_split": {
                        "reasoning_pct": 89.6,
                        "content_pct": 8.4,
                        "tool_call_pct": 2.0
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Google",
                "harness": "Mini-SWE-agent"
            },
            "google/gemini-3.5-flash": {
                "accuracy": 5.882,
                "latency": null,
                "stderr": 5.882,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 95.3,
                    "avg_query_seconds": 7.4,
                    "median_query_seconds": 4.6,
                    "avg_ttft_seconds": 3.0,
                    "output_time_split": {
                        "reasoning_pct": 53.8,
                        "content_pct": 45.5,
                        "tool_call_pct": 0.7
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Google",
                "harness": "Mini-SWE-agent"
            },
            "tencent/hy4-preview": {
                "accuracy": 5.882,
                "latency": null,
                "stderr": 5.882,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 178.9,
                    "avg_query_seconds": 26.0,
                    "median_query_seconds": 9.9,
                    "avg_ttft_seconds": 5.2,
                    "output_time_split": {
                        "reasoning_pct": 57.4,
                        "content_pct": 42.6,
                        "tool_call_pct": 0.0
                    }
                },
                "temperature": 0.9,
                "top_p": 1.0,
                "max_output_tokens": 64000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Tencent",
                "harness": "Terminus 2"
            },
            "xiaomi/mimo-v2.6-flash": {
                "accuracy": 5.882,
                "latency": null,
                "stderr": 5.882,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 129.1,
                    "avg_query_seconds": 55.0,
                    "median_query_seconds": 29.2,
                    "avg_ttft_seconds": 14.6,
                    "output_time_split": {
                        "reasoning_pct": 86.7,
                        "content_pct": 0.8,
                        "tool_call_pct": 12.5
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Xiaomi",
                "harness": "Mini-SWE-agent"
            },
            "zai/glm-5.3": {
                "accuracy": 5.882,
                "latency": null,
                "stderr": 5.882,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 236.0,
                    "avg_query_seconds": 27.0,
                    "median_query_seconds": 12.6,
                    "avg_ttft_seconds": 7.0,
                    "output_time_split": {
                        "reasoning_pct": 99.3,
                        "content_pct": 0.7,
                        "tool_call_pct": 0.0
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Fireworks AI",
                "harness": "Mini-SWE-agent"
            },
            "zai/glm-5.3-flash": {
                "accuracy": 5.882,
                "latency": null,
                "stderr": 5.882,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 153.3,
                    "avg_query_seconds": 39.4,
                    "median_query_seconds": 21.6,
                    "avg_ttft_seconds": 6.5,
                    "output_time_split": {
                        "reasoning_pct": 99.1,
                        "content_pct": 0.9,
                        "tool_call_pct": 0.0
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Zhipu AI",
                "harness": "Mini-SWE-agent"
            },
            "alibaba/qwen3.8-27b": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 476.3,
                    "avg_query_seconds": 28.2,
                    "median_query_seconds": 13.2,
                    "avg_ttft_seconds": 16.3,
                    "output_time_split": {
                        "reasoning_pct": 76.8,
                        "content_pct": 2.0,
                        "tool_call_pct": 21.2
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "xhigh",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Alibaba",
                "harness": "Mini-SWE-agent"
            },
            "deepseek/deepseek-v4.1-flash": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 156.4,
                    "avg_query_seconds": 10.1,
                    "median_query_seconds": 7.2,
                    "avg_ttft_seconds": 2.4,
                    "output_time_split": {
                        "reasoning_pct": 82.1,
                        "content_pct": 0.5,
                        "tool_call_pct": 17.4
                    }
                },
                "temperature": 1.0,
                "top_p": 1.0,
                "max_output_tokens": 384000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "DeepSeek",
                "harness": "Mini-SWE-agent"
            },
            "google/gemini-3.6-flash": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 16,
                    "avg_turns": 119.9,
                    "avg_query_seconds": 7.8,
                    "median_query_seconds": 5.3,
                    "avg_ttft_seconds": 3.1,
                    "output_time_split": {
                        "reasoning_pct": 91.7,
                        "content_pct": 7.6,
                        "tool_call_pct": 0.7
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Google",
                "harness": "Mini-SWE-agent"
            },
            "inception/mercury-2.5": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 78.3,
                    "avg_query_seconds": 2.5,
                    "median_query_seconds": 1.7,
                    "avg_ttft_seconds": 1.5,
                    "output_time_split": null
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Inception",
                "harness": "Mini-SWE-agent"
            },
            "kimi/kimi-k3": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 264.1,
                    "avg_query_seconds": 52.3,
                    "median_query_seconds": 30.4,
                    "avg_ttft_seconds": 20.3,
                    "output_time_split": {
                        "reasoning_pct": 74.9,
                        "content_pct": 2.6,
                        "tool_call_pct": 22.5
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 262144,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Moonshot AI",
                "harness": "Mini-SWE-agent"
            },
            "meta/muse_spark_1_3": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 133.8,
                    "avg_query_seconds": 9.5,
                    "median_query_seconds": 6.1,
                    "avg_ttft_seconds": 5.2,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": "xhigh",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Meta",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-5.6-luna": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 127.2,
                    "avg_query_seconds": 9.9,
                    "median_query_seconds": 6.4,
                    "avg_ttft_seconds": 5.5,
                    "output_time_split": {
                        "reasoning_pct": 22.2,
                        "content_pct": 1.5,
                        "tool_call_pct": 76.3
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-6-luna": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 89.8,
                    "avg_query_seconds": 19.1,
                    "median_query_seconds": 10.0,
                    "avg_ttft_seconds": 7.5,
                    "output_time_split": {
                        "reasoning_pct": 47.5,
                        "content_pct": 5.2,
                        "tool_call_pct": 47.3
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "reflection/Beam-1.0": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "temperature": 0.7,
                "top_p": 0.99,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Reflection",
                "harness": "Mini-SWE-agent"
            },
            "xiaomi/mimo-v2.6-pro": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 202.1,
                    "avg_query_seconds": 46.2,
                    "median_query_seconds": 23.0,
                    "avg_ttft_seconds": 13.9,
                    "output_time_split": {
                        "reasoning_pct": 85.1,
                        "content_pct": 1.4,
                        "tool_call_pct": 13.5
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Xiaomi",
                "harness": "Mini-SWE-agent"
            }
        },
        "earth_sciences": {
            "google/gemini-4-argon": {
                "accuracy": 62.5,
                "latency": null,
                "stderr": 18.298,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 137.0,
                    "avg_query_seconds": 39.7,
                    "median_query_seconds": 30.1,
                    "avg_ttft_seconds": 14.3,
                    "output_time_split": {
                        "reasoning_pct": 42.6,
                        "content_pct": 57.0,
                        "tool_call_pct": 0.4
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 262144,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Google",
                "harness": "Mini-SWE-agent"
            },
            "meta/muse-wonder-full": {
                "accuracy": 62.5,
                "latency": null,
                "stderr": 18.298,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 222.1,
                    "avg_query_seconds": 50.7,
                    "median_query_seconds": 18.6,
                    "avg_ttft_seconds": 48.9,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Meta",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-6-astra": {
                "accuracy": 62.5,
                "latency": null,
                "stderr": 18.298,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 44.8,
                    "avg_query_seconds": 106.7,
                    "median_query_seconds": 67.0,
                    "avg_ttft_seconds": 87.7,
                    "output_time_split": {
                        "reasoning_pct": 4.0,
                        "content_pct": 26.2,
                        "tool_call_pct": 69.8
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-6.1-sol": {
                "accuracy": 62.5,
                "latency": null,
                "stderr": 18.298,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 46.8,
                    "avg_query_seconds": 139.4,
                    "median_query_seconds": 66.1,
                    "avg_ttft_seconds": 121.8,
                    "output_time_split": {
                        "reasoning_pct": 4.0,
                        "content_pct": 11.8,
                        "tool_call_pct": 84.2
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "meta/muse-wonder": {
                "accuracy": 50.0,
                "latency": null,
                "stderr": 18.898,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 252.0,
                    "avg_query_seconds": 11.1,
                    "median_query_seconds": 6.6,
                    "avg_ttft_seconds": 10.0,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Meta",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-opus-5-5": {
                "accuracy": 37.5,
                "latency": null,
                "stderr": 18.298,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 126.2,
                    "avg_query_seconds": 18.6,
                    "median_query_seconds": 11.8,
                    "avg_ttft_seconds": 17.6,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-fable-5": {
                "accuracy": 25.0,
                "latency": null,
                "stderr": 16.366,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 166.0,
                    "avg_query_seconds": 20.5,
                    "median_query_seconds": 10.5,
                    "avg_ttft_seconds": 17.9,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-fable-5-1": {
                "accuracy": 25.0,
                "latency": null,
                "stderr": 16.366,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 119.0,
                    "avg_query_seconds": 21.8,
                    "median_query_seconds": 14.2,
                    "avg_ttft_seconds": 20.4,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "google/gemini-3.5-flash": {
                "accuracy": 25.0,
                "latency": null,
                "stderr": 16.366,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 102.1,
                    "avg_query_seconds": 9.0,
                    "median_query_seconds": 5.1,
                    "avg_ttft_seconds": 3.3,
                    "output_time_split": {
                        "reasoning_pct": 48.4,
                        "content_pct": 51.0,
                        "tool_call_pct": 0.6
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Google",
                "harness": "Mini-SWE-agent"
            },
            "grok/grok-4.7": {
                "accuracy": 25.0,
                "latency": null,
                "stderr": 16.366,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 99.6,
                    "avg_query_seconds": 21.6,
                    "median_query_seconds": 14.8,
                    "avg_ttft_seconds": 3.3,
                    "output_time_split": {
                        "reasoning_pct": 95.7,
                        "content_pct": 4.3,
                        "tool_call_pct": 0.0
                    }
                },
                "temperature": 0.7,
                "top_p": 0.95,
                "max_output_tokens": null,
                "reasoning": null,
                "reasoning_effort": "xhigh",
                "verbosity": null,
                "compute_effort": null,
                "provider": "SpaceXAI",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-5.6-sol": {
                "accuracy": 25.0,
                "latency": null,
                "stderr": 16.366,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 83.2,
                    "avg_query_seconds": 16.4,
                    "median_query_seconds": 11.9,
                    "avg_ttft_seconds": 8.3,
                    "output_time_split": {
                        "reasoning_pct": 20.3,
                        "content_pct": 29.1,
                        "tool_call_pct": 50.6
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-6-sol": {
                "accuracy": 25.0,
                "latency": null,
                "stderr": 16.366,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 129.5,
                    "avg_query_seconds": 21.3,
                    "median_query_seconds": 13.3,
                    "avg_ttft_seconds": 11.4,
                    "output_time_split": {
                        "reasoning_pct": 17.5,
                        "content_pct": 5.0,
                        "tool_call_pct": 77.5
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "alibaba/qwen3.8-max": {
                "accuracy": 12.5,
                "latency": null,
                "stderr": 12.5,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 198.8,
                    "avg_query_seconds": 42.4,
                    "median_query_seconds": 29.1,
                    "avg_ttft_seconds": 11.2,
                    "output_time_split": {
                        "reasoning_pct": 72.7,
                        "content_pct": 0.3,
                        "tool_call_pct": 27.0
                    }
                },
                "temperature": 0.7,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Alibaba",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-opus-5": {
                "accuracy": 12.5,
                "latency": null,
                "stderr": 12.5,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 246.1,
                    "avg_query_seconds": 16.8,
                    "median_query_seconds": 9.1,
                    "avg_ttft_seconds": 16.2,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-sonnet-5-5": {
                "accuracy": 12.5,
                "latency": null,
                "stderr": 12.5,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 313.9,
                    "avg_query_seconds": 13.9,
                    "median_query_seconds": 9.3,
                    "avg_ttft_seconds": 13.7,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "fireworks/ember-1": {
                "accuracy": 12.5,
                "latency": null,
                "stderr": 12.5,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 245.9,
                    "avg_query_seconds": 22.2,
                    "median_query_seconds": 15.3,
                    "avg_ttft_seconds": 6.8,
                    "output_time_split": {
                        "reasoning_pct": 64.4,
                        "content_pct": 1.7,
                        "tool_call_pct": 33.9
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Fireworks AI",
                "harness": "Mini-SWE-agent"
            },
            "google/gemini-3.6-flash": {
                "accuracy": 12.5,
                "latency": null,
                "stderr": 12.5,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 103.1,
                    "avg_query_seconds": 8.8,
                    "median_query_seconds": 6.3,
                    "avg_ttft_seconds": 3.2,
                    "output_time_split": {
                        "reasoning_pct": 99.5,
                        "content_pct": 0.0,
                        "tool_call_pct": 0.5
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Google",
                "harness": "Mini-SWE-agent"
            },
            "grok/grok-4.6": {
                "accuracy": 12.5,
                "latency": null,
                "stderr": 12.5,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 68.0,
                    "avg_query_seconds": 20.8,
                    "median_query_seconds": 11.9,
                    "avg_ttft_seconds": 5.3,
                    "output_time_split": {
                        "reasoning_pct": 76.8,
                        "content_pct": 23.2,
                        "tool_call_pct": 0.0
                    }
                },
                "temperature": 0.7,
                "top_p": 0.95,
                "max_output_tokens": null,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "SpaceXAI",
                "harness": "Mini-SWE-agent"
            },
            "kimi/kimi-k3": {
                "accuracy": 12.5,
                "latency": null,
                "stderr": 12.5,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 235.0,
                    "avg_query_seconds": 52.4,
                    "median_query_seconds": 31.3,
                    "avg_ttft_seconds": 20.8,
                    "output_time_split": {
                        "reasoning_pct": 72.9,
                        "content_pct": 2.6,
                        "tool_call_pct": 24.5
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 262144,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Moonshot AI",
                "harness": "Mini-SWE-agent"
            },
            "kimi/kimi-k3.1-pro": {
                "accuracy": 12.5,
                "latency": null,
                "stderr": 12.5,
                "cost_per_test": null,
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 262144,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Moonshot AI",
                "harness": "Mini-SWE-agent"
            },
            "reflection/Beam-1.0": {
                "accuracy": 12.5,
                "latency": null,
                "stderr": 12.5,
                "cost_per_test": null,
                "temperature": 0.7,
                "top_p": 0.99,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Reflection",
                "harness": "Mini-SWE-agent"
            },
            "alibaba/qwen3.8-27b": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 408.1,
                    "avg_query_seconds": 23.1,
                    "median_query_seconds": 12.1,
                    "avg_ttft_seconds": 14.0,
                    "output_time_split": {
                        "reasoning_pct": 69.9,
                        "content_pct": 1.7,
                        "tool_call_pct": 28.4
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "xhigh",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Alibaba",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-haiku-5-5": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 174.8,
                    "avg_query_seconds": 8.3,
                    "median_query_seconds": 4.1,
                    "avg_ttft_seconds": 6.8,
                    "output_time_split": null
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-opus-4-8": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 108.8,
                    "avg_query_seconds": 32.1,
                    "median_query_seconds": 18.2,
                    "avg_ttft_seconds": 25.9,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-sonnet-5": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 275.1,
                    "avg_query_seconds": 16.8,
                    "median_query_seconds": 8.5,
                    "avg_ttft_seconds": 15.9,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "deepseek/deepseek-v4-flash-0731": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 255.6,
                    "avg_query_seconds": 6.7,
                    "median_query_seconds": 4.7,
                    "avg_ttft_seconds": 2.6,
                    "output_time_split": {
                        "reasoning_pct": 79.5,
                        "content_pct": 0.4,
                        "tool_call_pct": 20.0
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 384000,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "DeepSeek",
                "harness": "Mini-SWE-agent"
            },
            "deepseek/deepseek-v4-pro-0813": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 286.4,
                    "avg_query_seconds": 9.1,
                    "median_query_seconds": 6.4,
                    "avg_ttft_seconds": 3.1,
                    "output_time_split": {
                        "reasoning_pct": 65.5,
                        "content_pct": 0.4,
                        "tool_call_pct": 34.1
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 384000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "DeepSeek",
                "harness": "Mini-SWE-agent"
            },
            "deepseek/deepseek-v4.1-flash": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 123.9,
                    "avg_query_seconds": 8.4,
                    "median_query_seconds": 5.9,
                    "avg_ttft_seconds": 2.0,
                    "output_time_split": {
                        "reasoning_pct": 79.8,
                        "content_pct": 0.3,
                        "tool_call_pct": 19.9
                    }
                },
                "temperature": 1.0,
                "top_p": 1.0,
                "max_output_tokens": 384000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "DeepSeek",
                "harness": "Mini-SWE-agent"
            },
            "google/gemini-3.1-pro-preview": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 152.2,
                    "avg_query_seconds": 9.5,
                    "median_query_seconds": 6.7,
                    "avg_ttft_seconds": 7.7,
                    "output_time_split": {
                        "reasoning_pct": 96.3,
                        "content_pct": 2.6,
                        "tool_call_pct": 1.2
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Google",
                "harness": "Mini-SWE-agent"
            },
            "google/gemini-3.7-flash": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 104.1,
                    "avg_query_seconds": 6.7,
                    "median_query_seconds": 5.1,
                    "avg_ttft_seconds": 4.3,
                    "output_time_split": {
                        "reasoning_pct": 98.5,
                        "content_pct": 0.0,
                        "tool_call_pct": 1.5
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Google",
                "harness": "Mini-SWE-agent"
            },
            "google/gemini-3.8-flash": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 136.2,
                    "avg_query_seconds": 5.4,
                    "median_query_seconds": 4.1,
                    "avg_ttft_seconds": 3.0,
                    "output_time_split": {
                        "reasoning_pct": 98.0,
                        "content_pct": 0.5,
                        "tool_call_pct": 1.5
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Google",
                "harness": "Mini-SWE-agent"
            },
            "inception/mercury-2.5": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 95.4,
                    "avg_query_seconds": 2.2,
                    "median_query_seconds": 1.9,
                    "avg_ttft_seconds": 1.5,
                    "output_time_split": null
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Inception",
                "harness": "Mini-SWE-agent"
            },
            "meta/muse_spark_1_3": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 147.1,
                    "avg_query_seconds": 8.7,
                    "median_query_seconds": 5.6,
                    "avg_ttft_seconds": 5.0,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": "xhigh",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Meta",
                "harness": "Mini-SWE-agent"
            },
            "meta/muse_spark_1_3_max": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 147.4,
                    "avg_query_seconds": 13.6,
                    "median_query_seconds": 9.4,
                    "avg_ttft_seconds": 6.1,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Meta",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-5.6-luna": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 131.1,
                    "avg_query_seconds": 8.9,
                    "median_query_seconds": 6.2,
                    "avg_ttft_seconds": 4.9,
                    "output_time_split": {
                        "reasoning_pct": 31.4,
                        "content_pct": 2.2,
                        "tool_call_pct": 66.4
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-5.6-terra": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 90.5,
                    "avg_query_seconds": 17.1,
                    "median_query_seconds": 10.4,
                    "avg_ttft_seconds": 7.3,
                    "output_time_split": {
                        "reasoning_pct": 28.5,
                        "content_pct": 10.4,
                        "tool_call_pct": 61.2
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-6-luna": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 70.5,
                    "avg_query_seconds": 20.5,
                    "median_query_seconds": 13.9,
                    "avg_ttft_seconds": 6.7,
                    "output_time_split": {
                        "reasoning_pct": 57.3,
                        "content_pct": 5.6,
                        "tool_call_pct": 37.0
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "stepfun/step-5-preview": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 253.9,
                    "avg_query_seconds": 19.3,
                    "median_query_seconds": 11.5,
                    "avg_ttft_seconds": 4.8,
                    "output_time_split": {
                        "reasoning_pct": 75.9,
                        "content_pct": 0.3,
                        "tool_call_pct": 23.8
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 1024000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Stepfun",
                "harness": "Mini-SWE-agent"
            },
            "tencent/hy4-preview": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 146.4,
                    "avg_query_seconds": 26.8,
                    "median_query_seconds": 11.9,
                    "avg_ttft_seconds": 4.5,
                    "output_time_split": {
                        "reasoning_pct": 55.8,
                        "content_pct": 44.2,
                        "tool_call_pct": 0.0
                    }
                },
                "temperature": 0.9,
                "top_p": 1.0,
                "max_output_tokens": 64000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Tencent",
                "harness": "Terminus 2"
            },
            "xiaomi/mimo-v2.6-flash": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 161.4,
                    "avg_query_seconds": 42.1,
                    "median_query_seconds": 27.4,
                    "avg_ttft_seconds": 11.8,
                    "output_time_split": {
                        "reasoning_pct": 83.5,
                        "content_pct": 0.8,
                        "tool_call_pct": 15.8
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Xiaomi",
                "harness": "Mini-SWE-agent"
            },
            "xiaomi/mimo-v2.6-pro": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 239.2,
                    "avg_query_seconds": 39.1,
                    "median_query_seconds": 22.0,
                    "avg_ttft_seconds": 12.8,
                    "output_time_split": {
                        "reasoning_pct": 83.0,
                        "content_pct": 1.5,
                        "tool_call_pct": 15.5
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Xiaomi",
                "harness": "Mini-SWE-agent"
            },
            "zai/glm-5.3": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 212.5,
                    "avg_query_seconds": 22.5,
                    "median_query_seconds": 12.2,
                    "avg_ttft_seconds": 6.6,
                    "output_time_split": {
                        "reasoning_pct": 99.4,
                        "content_pct": 0.6,
                        "tool_call_pct": 0.0
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Fireworks AI",
                "harness": "Mini-SWE-agent"
            },
            "zai/glm-5.3-flash": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 150.6,
                    "avg_query_seconds": 32.9,
                    "median_query_seconds": 19.6,
                    "avg_ttft_seconds": 6.2,
                    "output_time_split": {
                        "reasoning_pct": 98.7,
                        "content_pct": 1.2,
                        "tool_call_pct": 0.0
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Zhipu AI",
                "harness": "Mini-SWE-agent"
            }
        },
        "mathematical_sciences": {
            "openai/gpt-6-astra": {
                "accuracy": 88.235,
                "latency": null,
                "stderr": 8.055,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 51.9,
                    "avg_query_seconds": 88.0,
                    "median_query_seconds": 52.5,
                    "avg_ttft_seconds": 66.0,
                    "output_time_split": {
                        "reasoning_pct": 2.7,
                        "content_pct": 24.0,
                        "tool_call_pct": 73.3
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-6.1-sol": {
                "accuracy": 76.471,
                "latency": null,
                "stderr": 10.605,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 44.6,
                    "avg_query_seconds": 121.0,
                    "median_query_seconds": 58.8,
                    "avg_ttft_seconds": 93.9,
                    "output_time_split": {
                        "reasoning_pct": 4.6,
                        "content_pct": 10.1,
                        "tool_call_pct": 85.3
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-fable-5-1": {
                "accuracy": 64.706,
                "latency": null,
                "stderr": 11.947,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 96.5,
                    "avg_query_seconds": 32.3,
                    "median_query_seconds": 18.1,
                    "avg_ttft_seconds": 30.4,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-opus-5-5": {
                "accuracy": 58.824,
                "latency": null,
                "stderr": 12.304,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 102.3,
                    "avg_query_seconds": 26.1,
                    "median_query_seconds": 14.1,
                    "avg_ttft_seconds": 24.6,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-sonnet-5-5": {
                "accuracy": 58.824,
                "latency": null,
                "stderr": 12.304,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 287.2,
                    "avg_query_seconds": 16.1,
                    "median_query_seconds": 9.8,
                    "avg_ttft_seconds": 15.9,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "google/gemini-4-argon": {
                "accuracy": 58.824,
                "latency": null,
                "stderr": 12.304,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 110.2,
                    "avg_query_seconds": 58.6,
                    "median_query_seconds": 41.5,
                    "avg_ttft_seconds": 16.4,
                    "output_time_split": {
                        "reasoning_pct": 54.0,
                        "content_pct": 45.8,
                        "tool_call_pct": 0.3
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 262144,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Google",
                "harness": "Mini-SWE-agent"
            },
            "meta/muse-wonder-full": {
                "accuracy": 41.176,
                "latency": null,
                "stderr": 12.304,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 207.1,
                    "avg_query_seconds": 56.4,
                    "median_query_seconds": 16.5,
                    "avg_ttft_seconds": 54.6,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Meta",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-opus-5": {
                "accuracy": 29.412,
                "latency": null,
                "stderr": 11.391,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 176.4,
                    "avg_query_seconds": 21.5,
                    "median_query_seconds": 8.5,
                    "avg_ttft_seconds": 19.9,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-6-sol": {
                "accuracy": 29.412,
                "latency": null,
                "stderr": 11.391,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 125.8,
                    "avg_query_seconds": 23.0,
                    "median_query_seconds": 13.4,
                    "avg_ttft_seconds": 13.1,
                    "output_time_split": {
                        "reasoning_pct": 15.2,
                        "content_pct": 5.0,
                        "tool_call_pct": 79.8
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "kimi/kimi-k3.1-pro": {
                "accuracy": 23.529,
                "latency": null,
                "stderr": 10.605,
                "cost_per_test": null,
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 262144,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Moonshot AI",
                "harness": "Mini-SWE-agent"
            },
            "meta/muse-wonder": {
                "accuracy": 23.529,
                "latency": null,
                "stderr": 10.605,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 213.8,
                    "avg_query_seconds": 16.9,
                    "median_query_seconds": 7.9,
                    "avg_ttft_seconds": 15.6,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Meta",
                "harness": "Mini-SWE-agent"
            },
            "alibaba/qwen3.8-max": {
                "accuracy": 11.765,
                "latency": null,
                "stderr": 8.055,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 177.4,
                    "avg_query_seconds": 48.5,
                    "median_query_seconds": 28.4,
                    "avg_ttft_seconds": 11.1,
                    "output_time_split": {
                        "reasoning_pct": 76.7,
                        "content_pct": 0.4,
                        "tool_call_pct": 22.9
                    }
                },
                "temperature": 0.7,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Alibaba",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-fable-5": {
                "accuracy": 11.765,
                "latency": null,
                "stderr": 8.055,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 177.5,
                    "avg_query_seconds": 25.4,
                    "median_query_seconds": 8.8,
                    "avg_ttft_seconds": 23.1,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-haiku-5-5": {
                "accuracy": 11.765,
                "latency": null,
                "stderr": 8.055,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 226.0,
                    "avg_query_seconds": 9.3,
                    "median_query_seconds": 3.7,
                    "avg_ttft_seconds": 7.9,
                    "output_time_split": null
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "google/gemini-3.8-flash": {
                "accuracy": 11.765,
                "latency": null,
                "stderr": 8.055,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 149.1,
                    "avg_query_seconds": 7.0,
                    "median_query_seconds": 4.6,
                    "avg_ttft_seconds": 3.8,
                    "output_time_split": {
                        "reasoning_pct": 96.3,
                        "content_pct": 2.6,
                        "tool_call_pct": 1.1
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Google",
                "harness": "Mini-SWE-agent"
            },
            "meta/muse_spark_1_3_max": {
                "accuracy": 11.765,
                "latency": null,
                "stderr": 8.055,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 131.5,
                    "avg_query_seconds": 14.1,
                    "median_query_seconds": 8.1,
                    "avg_ttft_seconds": 6.0,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Meta",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-5.6-sol": {
                "accuracy": 11.765,
                "latency": null,
                "stderr": 8.055,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 93.9,
                    "avg_query_seconds": 21.1,
                    "median_query_seconds": 12.7,
                    "avg_ttft_seconds": 11.1,
                    "output_time_split": {
                        "reasoning_pct": 14.1,
                        "content_pct": 29.8,
                        "tool_call_pct": 56.1
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-5.6-terra": {
                "accuracy": 11.765,
                "latency": null,
                "stderr": 8.055,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 128.4,
                    "avg_query_seconds": 19.8,
                    "median_query_seconds": 11.1,
                    "avg_ttft_seconds": 10.2,
                    "output_time_split": {
                        "reasoning_pct": 24.3,
                        "content_pct": 9.6,
                        "tool_call_pct": 66.0
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-opus-4-8": {
                "accuracy": 5.882,
                "latency": null,
                "stderr": 5.882,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 101.9,
                    "avg_query_seconds": 46.0,
                    "median_query_seconds": 26.3,
                    "avg_ttft_seconds": 38.7,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "deepseek/deepseek-v4-flash-0731": {
                "accuracy": 5.882,
                "latency": null,
                "stderr": 5.882,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 206.8,
                    "avg_query_seconds": 8.4,
                    "median_query_seconds": 5.0,
                    "avg_ttft_seconds": 2.7,
                    "output_time_split": {
                        "reasoning_pct": 82.9,
                        "content_pct": 0.5,
                        "tool_call_pct": 16.6
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 384000,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "DeepSeek",
                "harness": "Mini-SWE-agent"
            },
            "deepseek/deepseek-v4-pro-0813": {
                "accuracy": 5.882,
                "latency": null,
                "stderr": 5.882,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 367.6,
                    "avg_query_seconds": 9.1,
                    "median_query_seconds": 5.5,
                    "avg_ttft_seconds": 2.9,
                    "output_time_split": {
                        "reasoning_pct": 69.8,
                        "content_pct": 0.5,
                        "tool_call_pct": 29.7
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 384000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "DeepSeek",
                "harness": "Mini-SWE-agent"
            },
            "fireworks/ember-1": {
                "accuracy": 5.882,
                "latency": null,
                "stderr": 5.882,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 308.3,
                    "avg_query_seconds": 26.9,
                    "median_query_seconds": 14.2,
                    "avg_ttft_seconds": 6.5,
                    "output_time_split": {
                        "reasoning_pct": 75.6,
                        "content_pct": 1.7,
                        "tool_call_pct": 22.7
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Fireworks AI",
                "harness": "Mini-SWE-agent"
            },
            "xiaomi/mimo-v2.6-flash": {
                "accuracy": 5.882,
                "latency": null,
                "stderr": 5.882,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 136.3,
                    "avg_query_seconds": 57.3,
                    "median_query_seconds": 26.8,
                    "avg_ttft_seconds": 15.1,
                    "output_time_split": {
                        "reasoning_pct": 88.3,
                        "content_pct": 0.6,
                        "tool_call_pct": 11.1
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Xiaomi",
                "harness": "Mini-SWE-agent"
            },
            "alibaba/qwen3.8-27b": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 351.8,
                    "avg_query_seconds": 33.0,
                    "median_query_seconds": 14.4,
                    "avg_ttft_seconds": 17.9,
                    "output_time_split": {
                        "reasoning_pct": 77.6,
                        "content_pct": 2.2,
                        "tool_call_pct": 20.2
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "xhigh",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Alibaba",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-sonnet-5": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 334.2,
                    "avg_query_seconds": 19.3,
                    "median_query_seconds": 8.4,
                    "avg_ttft_seconds": 18.5,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "deepseek/deepseek-v4.1-flash": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 141.3,
                    "avg_query_seconds": 10.8,
                    "median_query_seconds": 7.1,
                    "avg_ttft_seconds": 2.3,
                    "output_time_split": {
                        "reasoning_pct": 84.6,
                        "content_pct": 0.4,
                        "tool_call_pct": 15.0
                    }
                },
                "temperature": 1.0,
                "top_p": 1.0,
                "max_output_tokens": 384000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "DeepSeek",
                "harness": "Mini-SWE-agent"
            },
            "google/gemini-3.1-pro-preview": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 81.0,
                    "avg_query_seconds": 14.6,
                    "median_query_seconds": 8.4,
                    "avg_ttft_seconds": 10.7,
                    "output_time_split": {
                        "reasoning_pct": 86.3,
                        "content_pct": 13.3,
                        "tool_call_pct": 0.4
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Google",
                "harness": "Mini-SWE-agent"
            },
            "google/gemini-3.5-flash": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 88.4,
                    "avg_query_seconds": 10.6,
                    "median_query_seconds": 5.7,
                    "avg_ttft_seconds": 3.3,
                    "output_time_split": {
                        "reasoning_pct": 61.7,
                        "content_pct": 37.8,
                        "tool_call_pct": 0.5
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Google",
                "harness": "Mini-SWE-agent"
            },
            "google/gemini-3.6-flash": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 113.4,
                    "avg_query_seconds": 13.3,
                    "median_query_seconds": 9.8,
                    "avg_ttft_seconds": 3.9,
                    "output_time_split": {
                        "reasoning_pct": 87.6,
                        "content_pct": 11.8,
                        "tool_call_pct": 0.6
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Google",
                "harness": "Mini-SWE-agent"
            },
            "google/gemini-3.7-flash": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 130.1,
                    "avg_query_seconds": 8.0,
                    "median_query_seconds": 5.9,
                    "avg_ttft_seconds": 4.7,
                    "output_time_split": {
                        "reasoning_pct": 98.6,
                        "content_pct": 0.0,
                        "tool_call_pct": 1.4
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Google",
                "harness": "Mini-SWE-agent"
            },
            "grok/grok-4.6": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 93.6,
                    "avg_query_seconds": 22.5,
                    "median_query_seconds": 11.8,
                    "avg_ttft_seconds": 5.0,
                    "output_time_split": {
                        "reasoning_pct": 83.1,
                        "content_pct": 16.9,
                        "tool_call_pct": 0.0
                    }
                },
                "temperature": 0.7,
                "top_p": 0.95,
                "max_output_tokens": null,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "SpaceXAI",
                "harness": "Mini-SWE-agent"
            },
            "grok/grok-4.7": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 116.4,
                    "avg_query_seconds": 27.4,
                    "median_query_seconds": 17.1,
                    "avg_ttft_seconds": 4.3,
                    "output_time_split": {
                        "reasoning_pct": 95.2,
                        "content_pct": 4.8,
                        "tool_call_pct": 0.0
                    }
                },
                "temperature": 0.7,
                "top_p": 0.95,
                "max_output_tokens": null,
                "reasoning": null,
                "reasoning_effort": "xhigh",
                "verbosity": null,
                "compute_effort": null,
                "provider": "SpaceXAI",
                "harness": "Mini-SWE-agent"
            },
            "inception/mercury-2.5": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 100.6,
                    "avg_query_seconds": 2.7,
                    "median_query_seconds": 2.0,
                    "avg_ttft_seconds": 1.8,
                    "output_time_split": null
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Inception",
                "harness": "Mini-SWE-agent"
            },
            "kimi/kimi-k3": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 220.8,
                    "avg_query_seconds": 69.5,
                    "median_query_seconds": 45.8,
                    "avg_ttft_seconds": 28.9,
                    "output_time_split": {
                        "reasoning_pct": 77.7,
                        "content_pct": 2.1,
                        "tool_call_pct": 20.2
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 262144,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Moonshot AI",
                "harness": "Mini-SWE-agent"
            },
            "meta/muse_spark_1_3": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 133.1,
                    "avg_query_seconds": 10.4,
                    "median_query_seconds": 7.0,
                    "avg_ttft_seconds": 5.2,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": "xhigh",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Meta",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-5.6-luna": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 124.1,
                    "avg_query_seconds": 11.8,
                    "median_query_seconds": 7.5,
                    "avg_ttft_seconds": 7.0,
                    "output_time_split": {
                        "reasoning_pct": 23.7,
                        "content_pct": 1.2,
                        "tool_call_pct": 75.2
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-6-luna": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 91.3,
                    "avg_query_seconds": 18.5,
                    "median_query_seconds": 8.2,
                    "avg_ttft_seconds": 7.0,
                    "output_time_split": {
                        "reasoning_pct": 44.9,
                        "content_pct": 4.0,
                        "tool_call_pct": 51.1
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "reflection/Beam-1.0": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "temperature": 0.7,
                "top_p": 0.99,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Reflection",
                "harness": "Mini-SWE-agent"
            },
            "stepfun/step-5-preview": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 220.5,
                    "avg_query_seconds": 25.8,
                    "median_query_seconds": 11.4,
                    "avg_ttft_seconds": 5.9,
                    "output_time_split": {
                        "reasoning_pct": 82.9,
                        "content_pct": 0.4,
                        "tool_call_pct": 16.7
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 1024000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Stepfun",
                "harness": "Mini-SWE-agent"
            },
            "tencent/hy4-preview": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 257.5,
                    "avg_query_seconds": 37.0,
                    "median_query_seconds": 10.5,
                    "avg_ttft_seconds": 6.6,
                    "output_time_split": {
                        "reasoning_pct": 73.9,
                        "content_pct": 26.1,
                        "tool_call_pct": 0.0
                    }
                },
                "temperature": 0.9,
                "top_p": 1.0,
                "max_output_tokens": 64000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Tencent",
                "harness": "Terminus 2"
            },
            "xiaomi/mimo-v2.6-pro": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 168.5,
                    "avg_query_seconds": 57.1,
                    "median_query_seconds": 26.5,
                    "avg_ttft_seconds": 15.8,
                    "output_time_split": {
                        "reasoning_pct": 86.5,
                        "content_pct": 1.3,
                        "tool_call_pct": 12.3
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Xiaomi",
                "harness": "Mini-SWE-agent"
            },
            "zai/glm-5.3": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 217.7,
                    "avg_query_seconds": 27.7,
                    "median_query_seconds": 11.3,
                    "avg_ttft_seconds": 7.3,
                    "output_time_split": {
                        "reasoning_pct": 98.7,
                        "content_pct": 1.3,
                        "tool_call_pct": 0.0
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Fireworks AI",
                "harness": "Mini-SWE-agent"
            },
            "zai/glm-5.3-flash": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 17,
                    "avg_turns": 132.6,
                    "avg_query_seconds": 48.9,
                    "median_query_seconds": 21.2,
                    "avg_ttft_seconds": 6.5,
                    "output_time_split": {
                        "reasoning_pct": 98.6,
                        "content_pct": 1.3,
                        "tool_call_pct": 0.0
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Zhipu AI",
                "harness": "Mini-SWE-agent"
            }
        },
        "engineering_sciences": {
            "anthropic/claude-fable-5-1": {
                "accuracy": 44.444,
                "latency": null,
                "stderr": 17.568,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 129.3,
                    "avg_query_seconds": 31.4,
                    "median_query_seconds": 20.0,
                    "avg_ttft_seconds": 29.8,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-sonnet-5-5": {
                "accuracy": 44.444,
                "latency": null,
                "stderr": 17.568,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 357.1,
                    "avg_query_seconds": 17.5,
                    "median_query_seconds": 11.4,
                    "avg_ttft_seconds": 17.3,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-6-astra": {
                "accuracy": 33.333,
                "latency": null,
                "stderr": 16.667,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 48.9,
                    "avg_query_seconds": 164.7,
                    "median_query_seconds": 103.5,
                    "avg_ttft_seconds": 134.7,
                    "output_time_split": {
                        "reasoning_pct": 1.7,
                        "content_pct": 12.3,
                        "tool_call_pct": 86.0
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-6.1-sol": {
                "accuracy": 33.333,
                "latency": null,
                "stderr": 16.667,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 46.6,
                    "avg_query_seconds": 168.0,
                    "median_query_seconds": 89.2,
                    "avg_ttft_seconds": 138.0,
                    "output_time_split": {
                        "reasoning_pct": 3.0,
                        "content_pct": 7.7,
                        "tool_call_pct": 89.3
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-opus-5-5": {
                "accuracy": 22.222,
                "latency": null,
                "stderr": 14.699,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 136.9,
                    "avg_query_seconds": 23.4,
                    "median_query_seconds": 15.4,
                    "avg_ttft_seconds": 22.4,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "kimi/kimi-k3.1-pro": {
                "accuracy": 22.222,
                "latency": null,
                "stderr": 14.699,
                "cost_per_test": null,
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 262144,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Moonshot AI",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-6-sol": {
                "accuracy": 22.222,
                "latency": null,
                "stderr": 14.699,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 129.7,
                    "avg_query_seconds": 30.8,
                    "median_query_seconds": 20.1,
                    "avg_ttft_seconds": 15.4,
                    "output_time_split": {
                        "reasoning_pct": 12.6,
                        "content_pct": 8.7,
                        "tool_call_pct": 78.7
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-opus-5": {
                "accuracy": 11.111,
                "latency": null,
                "stderr": 11.111,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 189.8,
                    "avg_query_seconds": 25.3,
                    "median_query_seconds": 13.7,
                    "avg_ttft_seconds": 24.2,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "google/gemini-3.8-flash": {
                "accuracy": 11.111,
                "latency": null,
                "stderr": 11.111,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 8,
                    "avg_turns": 149.4,
                    "avg_query_seconds": 5.9,
                    "median_query_seconds": 4.7,
                    "avg_ttft_seconds": 3.3,
                    "output_time_split": {
                        "reasoning_pct": 98.8,
                        "content_pct": 0.0,
                        "tool_call_pct": 1.2
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Google",
                "harness": "Mini-SWE-agent"
            },
            "google/gemini-4-argon": {
                "accuracy": 11.111,
                "latency": null,
                "stderr": 11.111,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 119.8,
                    "avg_query_seconds": 54.1,
                    "median_query_seconds": 41.7,
                    "avg_ttft_seconds": 16.2,
                    "output_time_split": {
                        "reasoning_pct": 53.5,
                        "content_pct": 46.2,
                        "tool_call_pct": 0.3
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 262144,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Google",
                "harness": "Mini-SWE-agent"
            },
            "grok/grok-4.6": {
                "accuracy": 11.111,
                "latency": null,
                "stderr": 11.111,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 48.6,
                    "avg_query_seconds": 33.7,
                    "median_query_seconds": 24.7,
                    "avg_ttft_seconds": 5.4,
                    "output_time_split": {
                        "reasoning_pct": 78.8,
                        "content_pct": 21.2,
                        "tool_call_pct": 0.0
                    }
                },
                "temperature": 0.7,
                "top_p": 0.95,
                "max_output_tokens": null,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "SpaceXAI",
                "harness": "Mini-SWE-agent"
            },
            "meta/muse-wonder-full": {
                "accuracy": 11.111,
                "latency": null,
                "stderr": 11.111,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 185.3,
                    "avg_query_seconds": 46.0,
                    "median_query_seconds": 21.8,
                    "avg_ttft_seconds": 43.5,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Meta",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-5.6-sol": {
                "accuracy": 11.111,
                "latency": null,
                "stderr": 11.111,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 79.1,
                    "avg_query_seconds": 18.8,
                    "median_query_seconds": 13.4,
                    "avg_ttft_seconds": 9.2,
                    "output_time_split": {
                        "reasoning_pct": 16.3,
                        "content_pct": 23.7,
                        "tool_call_pct": 60.0
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-6-luna": {
                "accuracy": 11.111,
                "latency": null,
                "stderr": 11.111,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 56.6,
                    "avg_query_seconds": 28.9,
                    "median_query_seconds": 17.2,
                    "avg_ttft_seconds": 7.4,
                    "output_time_split": {
                        "reasoning_pct": 39.7,
                        "content_pct": 3.1,
                        "tool_call_pct": 57.1
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "alibaba/qwen3.8-27b": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 321.6,
                    "avg_query_seconds": 42.0,
                    "median_query_seconds": 18.0,
                    "avg_ttft_seconds": 24.8,
                    "output_time_split": {
                        "reasoning_pct": 75.2,
                        "content_pct": 1.7,
                        "tool_call_pct": 23.1
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "xhigh",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Alibaba",
                "harness": "Mini-SWE-agent"
            },
            "alibaba/qwen3.8-max": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 204.4,
                    "avg_query_seconds": 56.1,
                    "median_query_seconds": 36.5,
                    "avg_ttft_seconds": 11.6,
                    "output_time_split": {
                        "reasoning_pct": 78.3,
                        "content_pct": 0.2,
                        "tool_call_pct": 21.5
                    }
                },
                "temperature": 0.7,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Alibaba",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-fable-5": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 172.4,
                    "avg_query_seconds": 31.6,
                    "median_query_seconds": 13.9,
                    "avg_ttft_seconds": 27.8,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-haiku-5-5": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 154.7,
                    "avg_query_seconds": 12.5,
                    "median_query_seconds": 6.9,
                    "avg_ttft_seconds": 10.4,
                    "output_time_split": null
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-opus-4-8": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 91.9,
                    "avg_query_seconds": 42.0,
                    "median_query_seconds": 30.9,
                    "avg_ttft_seconds": 32.7,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "anthropic/claude-sonnet-5": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 287.4,
                    "avg_query_seconds": 17.6,
                    "median_query_seconds": 8.3,
                    "avg_ttft_seconds": 16.5,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": "max",
                "provider": "Anthropic",
                "harness": "Mini-SWE-agent"
            },
            "deepseek/deepseek-v4-flash-0731": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 186.9,
                    "avg_query_seconds": 9.1,
                    "median_query_seconds": 6.7,
                    "avg_ttft_seconds": 2.7,
                    "output_time_split": {
                        "reasoning_pct": 78.6,
                        "content_pct": 0.5,
                        "tool_call_pct": 20.9
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 384000,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "DeepSeek",
                "harness": "Mini-SWE-agent"
            },
            "deepseek/deepseek-v4-pro-0813": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 282.6,
                    "avg_query_seconds": 12.0,
                    "median_query_seconds": 8.7,
                    "avg_ttft_seconds": 3.0,
                    "output_time_split": {
                        "reasoning_pct": 65.4,
                        "content_pct": 0.4,
                        "tool_call_pct": 34.2
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 384000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "DeepSeek",
                "harness": "Mini-SWE-agent"
            },
            "deepseek/deepseek-v4.1-flash": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 169.6,
                    "avg_query_seconds": 10.0,
                    "median_query_seconds": 7.9,
                    "avg_ttft_seconds": 2.4,
                    "output_time_split": {
                        "reasoning_pct": 78.4,
                        "content_pct": 0.2,
                        "tool_call_pct": 21.4
                    }
                },
                "temperature": 1.0,
                "top_p": 1.0,
                "max_output_tokens": 384000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "DeepSeek",
                "harness": "Mini-SWE-agent"
            },
            "fireworks/ember-1": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 334.8,
                    "avg_query_seconds": 24.4,
                    "median_query_seconds": 14.3,
                    "avg_ttft_seconds": 6.2,
                    "output_time_split": {
                        "reasoning_pct": 66.9,
                        "content_pct": 1.4,
                        "tool_call_pct": 31.7
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Fireworks AI",
                "harness": "Mini-SWE-agent"
            },
            "google/gemini-3.1-pro-preview": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 73.8,
                    "avg_query_seconds": 13.5,
                    "median_query_seconds": 8.2,
                    "avg_ttft_seconds": 9.5,
                    "output_time_split": {
                        "reasoning_pct": 70.5,
                        "content_pct": 29.1,
                        "tool_call_pct": 0.4
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Google",
                "harness": "Mini-SWE-agent"
            },
            "google/gemini-3.5-flash": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 61.1,
                    "avg_query_seconds": 10.9,
                    "median_query_seconds": 6.2,
                    "avg_ttft_seconds": 3.0,
                    "output_time_split": {
                        "reasoning_pct": 55.8,
                        "content_pct": 43.9,
                        "tool_call_pct": 0.3
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Google",
                "harness": "Mini-SWE-agent"
            },
            "google/gemini-3.6-flash": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 70.4,
                    "avg_query_seconds": 7.7,
                    "median_query_seconds": 6.1,
                    "avg_ttft_seconds": 2.9,
                    "output_time_split": {
                        "reasoning_pct": 99.3,
                        "content_pct": 0.0,
                        "tool_call_pct": 0.7
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Google",
                "harness": "Mini-SWE-agent"
            },
            "google/gemini-3.7-flash": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 103.8,
                    "avg_query_seconds": 6.8,
                    "median_query_seconds": 5.3,
                    "avg_ttft_seconds": 4.2,
                    "output_time_split": {
                        "reasoning_pct": 86.8,
                        "content_pct": 11.7,
                        "tool_call_pct": 1.5
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Google",
                "harness": "Mini-SWE-agent"
            },
            "grok/grok-4.7": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 82.1,
                    "avg_query_seconds": 29.6,
                    "median_query_seconds": 23.7,
                    "avg_ttft_seconds": 3.0,
                    "output_time_split": {
                        "reasoning_pct": 92.8,
                        "content_pct": 7.2,
                        "tool_call_pct": 0.0
                    }
                },
                "temperature": 0.7,
                "top_p": 0.95,
                "max_output_tokens": null,
                "reasoning": null,
                "reasoning_effort": "xhigh",
                "verbosity": null,
                "compute_effort": null,
                "provider": "SpaceXAI",
                "harness": "Mini-SWE-agent"
            },
            "inception/mercury-2.5": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 50.0,
                    "avg_query_seconds": 2.6,
                    "median_query_seconds": 2.1,
                    "avg_ttft_seconds": 1.4,
                    "output_time_split": null
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 65536,
                "reasoning": null,
                "reasoning_effort": "high",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Inception",
                "harness": "Mini-SWE-agent"
            },
            "kimi/kimi-k3": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 244.8,
                    "avg_query_seconds": 66.1,
                    "median_query_seconds": 50.5,
                    "avg_ttft_seconds": 25.0,
                    "output_time_split": {
                        "reasoning_pct": 70.2,
                        "content_pct": 2.4,
                        "tool_call_pct": 27.4
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 262144,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Moonshot AI",
                "harness": "Mini-SWE-agent"
            },
            "meta/muse-wonder": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 227.0,
                    "avg_query_seconds": 16.2,
                    "median_query_seconds": 9.2,
                    "avg_ttft_seconds": 14.9,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Meta",
                "harness": "Mini-SWE-agent"
            },
            "meta/muse_spark_1_3": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 128.0,
                    "avg_query_seconds": 10.2,
                    "median_query_seconds": 7.2,
                    "avg_ttft_seconds": 5.3,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": "xhigh",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Meta",
                "harness": "Mini-SWE-agent"
            },
            "meta/muse_spark_1_3_max": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 128.3,
                    "avg_query_seconds": 16.0,
                    "median_query_seconds": 11.0,
                    "avg_ttft_seconds": 6.6,
                    "output_time_split": null
                },
                "temperature": 1.0,
                "top_p": null,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Meta",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-5.6-luna": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 125.0,
                    "avg_query_seconds": 10.6,
                    "median_query_seconds": 7.8,
                    "avg_ttft_seconds": 5.5,
                    "output_time_split": {
                        "reasoning_pct": 26.1,
                        "content_pct": 1.4,
                        "tool_call_pct": 72.5
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "openai/gpt-5.6-terra": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 98.3,
                    "avg_query_seconds": 19.8,
                    "median_query_seconds": 14.0,
                    "avg_ttft_seconds": 8.4,
                    "output_time_split": {
                        "reasoning_pct": 28.6,
                        "content_pct": 7.2,
                        "tool_call_pct": 64.2
                    }
                },
                "temperature": null,
                "top_p": null,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "OpenAI",
                "harness": "Mini-SWE-agent"
            },
            "reflection/Beam-1.0": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "temperature": 0.7,
                "top_p": 0.99,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Reflection",
                "harness": "Mini-SWE-agent"
            },
            "stepfun/step-5-preview": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 174.9,
                    "avg_query_seconds": 26.5,
                    "median_query_seconds": 14.3,
                    "avg_ttft_seconds": 4.9,
                    "output_time_split": {
                        "reasoning_pct": 79.9,
                        "content_pct": 0.6,
                        "tool_call_pct": 19.5
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 1024000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Stepfun",
                "harness": "Mini-SWE-agent"
            },
            "tencent/hy4-preview": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 157.8,
                    "avg_query_seconds": 35.6,
                    "median_query_seconds": 12.7,
                    "avg_ttft_seconds": 5.4,
                    "output_time_split": {
                        "reasoning_pct": 62.3,
                        "content_pct": 37.7,
                        "tool_call_pct": 0.0
                    }
                },
                "temperature": 0.9,
                "top_p": 1.0,
                "max_output_tokens": 64000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Tencent",
                "harness": "Terminus 2"
            },
            "xiaomi/mimo-v2.6-flash": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 126.7,
                    "avg_query_seconds": 53.7,
                    "median_query_seconds": 30.8,
                    "avg_ttft_seconds": 13.4,
                    "output_time_split": {
                        "reasoning_pct": 84.8,
                        "content_pct": 0.8,
                        "tool_call_pct": 14.4
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Xiaomi",
                "harness": "Mini-SWE-agent"
            },
            "xiaomi/mimo-v2.6-pro": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 133.6,
                    "avg_query_seconds": 59.7,
                    "median_query_seconds": 33.5,
                    "avg_ttft_seconds": 11.5,
                    "output_time_split": {
                        "reasoning_pct": 84.3,
                        "content_pct": 1.4,
                        "tool_call_pct": 14.3
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": null,
                "verbosity": null,
                "compute_effort": null,
                "provider": "Xiaomi",
                "harness": "Mini-SWE-agent"
            },
            "zai/glm-5.3": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 246.7,
                    "avg_query_seconds": 27.4,
                    "median_query_seconds": 13.8,
                    "avg_ttft_seconds": 6.9,
                    "output_time_split": {
                        "reasoning_pct": 99.1,
                        "content_pct": 0.9,
                        "tool_call_pct": 0.0
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 131072,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Fireworks AI",
                "harness": "Mini-SWE-agent"
            },
            "zai/glm-5.3-flash": {
                "accuracy": 0.0,
                "latency": null,
                "stderr": 0.0,
                "cost_per_test": null,
                "turn_metrics": {
                    "tasks_measured": 9,
                    "avg_turns": 166.0,
                    "avg_query_seconds": 43.5,
                    "median_query_seconds": 26.4,
                    "avg_ttft_seconds": 6.8,
                    "output_time_split": {
                        "reasoning_pct": 98.8,
                        "content_pct": 1.2,
                        "tool_call_pct": 0.0
                    }
                },
                "temperature": 1.0,
                "top_p": 0.95,
                "max_output_tokens": 128000,
                "reasoning": null,
                "reasoning_effort": "max",
                "verbosity": null,
                "compute_effort": null,
                "provider": "Zhipu AI",
                "harness": "Mini-SWE-agent"
            }
        }
    }
}