{
  "metadata": {
    "benchmark": "Vals Router Index",
    "slug": "vals_router_index",
    "description": "How model routers compare with single models on long agentic work from the Vals Index.",
    "benchmark_id": "vals_router_index",
    "family": "vals_router_index",
    "version": "1.0",
    "updated": "2026-10-20",
    "dataset_type": "private",
    "industry": "index",
    "tasks": {
      "overall": "Overall",
      "finance_agent": "Finance Agent",
      "emb": "EMB",
      "terminal_bench_4": "Terminal-Bench 4.0",
      "vibe_code_bench": "Vibe Code Bench",
      "code_migration": "Code Migration",
      "legal_research": "Legal Research",
      "legal_agent_benchmark": "HLAB",
      "tax_agent_bench": "Tax Agent Bench"
    },
    "models": [
      "fireworks/firerouter",
      "openrouter/openrouter/auto",
      "openrouter/typesafe/jev-router",
      "ramp/switchyard",
      "sakana/fugu-max-v1.0"
    ],
    "partners": [],
    "showBadge": false,
    "visible": true,
    "use_cost_per_test": true,
    "runner": "external",
    "mode": "agentic",
    "archived": false,
    "partner": false,
    "total_models": 5
  },
  "tasks": {
    "overall": {
      "openrouter/openrouter/auto": {
        "accuracy": 52.195,
        "latency": 2420.133,
        "stderr": 1.207,
        "cost_per_test": 0.817334,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      },
      "fireworks/firerouter": {
        "accuracy": 56.458,
        "latency": 2807.558,
        "stderr": 1.303,
        "cost_per_test": 9.217354,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      },
      "ramp/switchyard": {
        "accuracy": 57.795,
        "latency": 680.404,
        "stderr": 1.167,
        "cost_per_test": 0.956238,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      },
      "openrouter/typesafe/jev-router": {
        "accuracy": 55.652,
        "latency": 1478.241,
        "stderr": 1.332,
        "cost_per_test": 5.303181,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      },
      "sakana/fugu-max-v1.0": {
        "accuracy": 48.114,
        "latency": 850.32,
        "stderr": 1.166,
        "cost_per_test": 3.912391,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      }
    },
    "finance_agent": {
      "openrouter/openrouter/auto": {
        "accuracy": 59.098,
        "latency": 268.433,
        "stderr": 2.201,
        "cost_per_test": 0.09512,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      },
      "fireworks/firerouter": {
        "accuracy": 64.456,
        "latency": 147.97,
        "stderr": 2.14,
        "cost_per_test": 0.653802,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      },
      "ramp/switchyard": {
        "accuracy": 62.451,
        "latency": 127.076,
        "stderr": 2.145,
        "cost_per_test": 0.145253,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      },
      "openrouter/typesafe/jev-router": {
        "accuracy": 59.614,
        "latency": 261.498,
        "stderr": 2.183,
        "cost_per_test": 0.895878,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      },
      "sakana/fugu-max-v1.0": {
        "accuracy": 54.99,
        "latency": 177.443,
        "stderr": 2.217,
        "cost_per_test": 0.952185,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      }
    },
    "emb": {
      "openrouter/openrouter/auto": {
        "accuracy": 67.174,
        "latency": 902.021,
        "stderr": 1.95,
        "cost_per_test": 0.204939,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      },
      "fireworks/firerouter": {
        "accuracy": 73.016,
        "latency": 731.624,
        "stderr": 1.895,
        "cost_per_test": 2.002952,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      },
      "ramp/switchyard": {
        "accuracy": 69.61,
        "latency": 354.757,
        "stderr": 2.223,
        "cost_per_test": 0.378797,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      },
      "openrouter/typesafe/jev-router": {
        "accuracy": 70.977,
        "latency": 551.08,
        "stderr": 2.23,
        "cost_per_test": 2.029949,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      },
      "sakana/fugu-max-v1.0": {
        "accuracy": 66.927,
        "latency": 460.153,
        "stderr": 2.213,
        "cost_per_test": 0.854517,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      }
    },
    "terminal_bench_4": {
      "openrouter/openrouter/auto": {
        "accuracy": 25.758,
        "latency": 2604.973,
        "stderr": 4.058,
        "cost_per_test": 0.862076,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      },
      "fireworks/firerouter": {
        "accuracy": 46.465,
        "latency": 3465.113,
        "stderr": 5.38,
        "cost_per_test": 7.274694,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      },
      "ramp/switchyard": {
        "accuracy": 32.323,
        "latency": 1184.575,
        "stderr": 4.772,
        "cost_per_test": 1.089203,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      },
      "openrouter/typesafe/jev-router": {
        "accuracy": 45.96,
        "latency": 1353.451,
        "stderr": 5.178,
        "cost_per_test": 2.375726,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      },
      "sakana/fugu-max-v1.0": {
        "accuracy": 17.172,
        "latency": 1901.851,
        "stderr": 3.616,
        "cost_per_test": 8.364969,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      }
    },
    "vibe_code_bench": {
      "openrouter/openrouter/auto": {
        "accuracy": 72.257,
        "latency": 2642.737,
        "stderr": 4.666,
        "cost_per_test": 1.002777,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      },
      "fireworks/firerouter": {
        "accuracy": 78.786,
        "latency": 1205.389,
        "stderr": 4.649,
        "cost_per_test": 6.51642,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      },
      "ramp/switchyard": {
        "accuracy": 86.475,
        "latency": 1309.887,
        "stderr": 2.124,
        "cost_per_test": 3.16861,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      },
      "openrouter/typesafe/jev-router": {
        "accuracy": 61.391,
        "latency": 856.124,
        "stderr": 5.535,
        "cost_per_test": 4.551245,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      },
      "sakana/fugu-max-v1.0": {
        "accuracy": 69.608,
        "latency": 1750.313,
        "stderr": 4.14,
        "cost_per_test": 15.437519,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      }
    },
    "code_migration": {
      "openrouter/openrouter/auto": {
        "accuracy": 44.815,
        "latency": 10049.686,
        "stderr": 4.266,
        "cost_per_test": 3.736539,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      },
      "fireworks/firerouter": {
        "accuracy": 21.25,
        "latency": 14502.675,
        "stderr": 4.871,
        "cost_per_test": 47.473857,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      },
      "ramp/switchyard": {
        "accuracy": 52.61,
        "latency": 1451.158,
        "stderr": 4.261,
        "cost_per_test": 1.215162,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      },
      "openrouter/typesafe/jev-router": {
        "accuracy": 54.683,
        "latency": 6605.332,
        "stderr": 3.918,
        "cost_per_test": 26.506681,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      },
      "sakana/fugu-max-v1.0": {
        "accuracy": 33.946,
        "latency": 1391.26,
        "stderr": 3.816,
        "cost_per_test": 2.702462,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      }
    },
    "legal_research": {
      "openrouter/openrouter/auto": {
        "accuracy": 29.808,
        "latency": 1808.769,
        "stderr": 3.179,
        "cost_per_test": 0.207931,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      },
      "fireworks/firerouter": {
        "accuracy": 41.346,
        "latency": 383.829,
        "stderr": 3.423,
        "cost_per_test": 0.727174,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      },
      "ramp/switchyard": {
        "accuracy": 36.538,
        "latency": 236.513,
        "stderr": 3.347,
        "cost_per_test": 0.258068,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      },
      "openrouter/typesafe/jev-router": {
        "accuracy": 29.808,
        "latency": 415.838,
        "stderr": 3.179,
        "cost_per_test": 0.902983,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      },
      "sakana/fugu-max-v1.0": {
        "accuracy": 24.519,
        "latency": 373.239,
        "stderr": 2.99,
        "cost_per_test": 0.713539,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      }
    },
    "legal_agent_benchmark": {
      "openrouter/openrouter/auto": {
        "accuracy": 5.0,
        "latency": 713.15,
        "stderr": 1.814,
        "cost_per_test": 0.395779,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      },
      "fireworks/firerouter": {
        "accuracy": 6.25,
        "latency": 1850.28,
        "stderr": 2.012,
        "cost_per_test": 8.739914,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      },
      "ramp/switchyard": {
        "accuracy": 5.0,
        "latency": 629.575,
        "stderr": 1.908,
        "cost_per_test": 1.251771,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      },
      "openrouter/typesafe/jev-router": {
        "accuracy": 1.25,
        "latency": 1466.576,
        "stderr": 0.929,
        "cost_per_test": 4.470621,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      },
      "sakana/fugu-max-v1.0": {
        "accuracy": 2.917,
        "latency": 486.197,
        "stderr": 1.226,
        "cost_per_test": 1.779858,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      }
    },
    "tax_agent_bench": {
      "openrouter/openrouter/auto": {
        "accuracy": 11.996,
        "latency": 371.294,
        "stderr": 2.308,
        "cost_per_test": 0.03351,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      },
      "fireworks/firerouter": {
        "accuracy": 23.796,
        "latency": 173.584,
        "stderr": 3.07,
        "cost_per_test": 0.350018,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      },
      "ramp/switchyard": {
        "accuracy": 22.28,
        "latency": 149.691,
        "stderr": 3.003,
        "cost_per_test": 0.143038,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      },
      "openrouter/typesafe/jev-router": {
        "accuracy": 16.036,
        "latency": 316.031,
        "stderr": 2.646,
        "cost_per_test": 0.692363,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      },
      "sakana/fugu-max-v1.0": {
        "accuracy": 13.316,
        "latency": 262.102,
        "stderr": 2.44,
        "cost_per_test": 0.494077,
        "temperature": null,
        "top_p": null,
        "max_output_tokens": null,
        "reasoning": null,
        "reasoning_effort": null,
        "verbosity": null,
        "compute_effort": null,
        "provider": null,
        "harness": null
      }
    }
  }
}
