Model comparison
Gemini 3.7 Flash vs MiMo V2.6 Pro: Benchmark Comparison
Gemini 3.7 Flash has the higher score on 7 of 18 shared benchmarks; MiMo V2.6 Pro leads on 11.
The largest observed score gap is 29.11 pts on CyberBench v1.1 , where MiMo V2.6 Pro leads.
Reported ±1 standard-error ranges overlap on 7 of 18 shared benchmarks with comparable uncertainty data. This is not a pairwise statistical significance test.
Shared benchmark results
Scores and reported standard errors come from the latest visible Vals benchmark version available for both models.
| Benchmark | Gemini 3.7 Flash | MiMo V2.6 Pro | Gap | Reported uncertainty |
|---|---|---|---|---|
| Vals Index | 59.31% ±1.06 | 59.47% ±1.18 | 0.16 pts | Reported ±1 SE ranges overlap |
| Harvey's Legal Agent Benchmark | 8.75% ±2.15 | 10.83% ±2.17 | 2.08 pts | Reported ±1 SE ranges overlap |
| Legal Research Bench | 34.62% ±3.31 | 47.12% ±3.47 | 12.50 pts | Reported ±1 SE ranges do not overlap |
| EMB | 71.33% ±2.26 | 62.86% ±3.14 | 8.47 pts | Reported ±1 SE ranges do not overlap |
| Finance Agent (v2) | 59.04% ±0.27 | 57.34% ±0.57 | 1.70 pts | Reported ±1 SE ranges do not overlap |
| Tax Agent Bench | 57.66% ±3.31 | 64.94% ±3.21 | 7.27 pts | Reported ±1 SE ranges do not overlap |
| MedCode | 53.39% ±2.12 | 44.97% ±2.10 | 8.42 pts | Reported ±1 SE ranges do not overlap |
| MedScribe | 83.94% ±2.00 | 88.31% ±1.94 | 4.37 pts | Reported ±1 SE ranges do not overlap |
| ProofBench v1.1 | 58.00% ±4.96 | 70.00% ±4.61 | 12.00 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench Science | 5.71% ±2.79 | 2.86% ±2.01 | 2.86 pts | Reported ±1 SE ranges overlap |
| SAGE | 49.23% ±3.38 | 45.05% ±3.40 | 4.18 pts | Reported ±1 SE ranges overlap |
| Code Migration | 34.80% ±4.22 | 43.01% ±4.32 | 8.21 pts | Reported ±1 SE ranges overlap |
| IOI | 67.83% ±3.61 | 39.33% ±2.41 | 28.50 pts | Reported ±1 SE ranges do not overlap |
| ProgramBench | 0.00% ±0.00 | 0.50% ±0.50 | 0.50 pts | Reported ±1 SE ranges overlap |
| Terminal-Bench 4.0 | 6.06% ±0.88 | 24.75% ±3.07 | 18.69 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench v1.1 | 70.39% ±4.84 | 85.22% ±3.39 | 14.83 pts | Reported ±1 SE ranges do not overlap |
| CyberBench v1.1 | 43.75% ±2.21 | 72.86% ±5.50 | 29.11 pts | Reported ±1 SE ranges do not overlap |
| SRE Bench | 4.58% ±1.29 | 3.05% ±1.06 | 1.53 pts | Reported ±1 SE ranges overlap |
Performance by category
| Category | Gemini 3.7 Flash average | MiMo V2.6 Pro average |
|---|---|---|
| Index | 59.31% | 59.47% |
| Legal | 21.68% | 28.97% |
| Finance | 62.68% | 61.71% |
| Healthcare | 68.67% | 66.64% |
| Math | 58.00% | 70.00% |
| Science | 5.71% | 2.86% |
| Education | 49.23% | 45.05% |
| Coding | 35.82% | 38.56% |
| Beta | 24.16% | 37.95% |
Cost and latency
Cost per test appears only where the benchmark reports it for both models. Latency is the measured completion time for that benchmark.
| Benchmark | Gemini 3.7 Flash cost | MiMo V2.6 Pro cost | Gemini 3.7 Flash latency | MiMo V2.6 Pro latency |
|---|---|---|---|---|
| Vals Index | $4.17 | $0.39 | 43m13s | 52m20s |
| Harvey's Legal Agent Benchmark | $2.56 | $0.22 | 7m32s | 25m49s |
| Legal Research Bench | $0.76 | $0.18 | 4m49s | 30m16s |
| EMB | $5.69 | $0.33 | 10m51s | 55m14s |
| Finance Agent (v2) | $1.48 | $0.20 | 3m01s | 10m48s |
| Tax Agent Bench | $0.48 | $0.13 | 87.24s | 25m48s |
| MedCode | N/A | N/A | 9.16s | 2m57s |
| MedScribe | N/A | N/A | 16.82s | 2m53s |
| ProofBench v1.1 | $0.56 | $0.25 | 7m04s | 1h02m |
| Terminal-Bench Science | $13.22 | $0.68 | 1h08m | 3h40m |
| SAGE | N/A | N/A | 9.79s | 3m40s |
| Code Migration | $21.46 | $0.69 | 2h39m | 2h49m |
| IOI | $3.40 | $0.84 | 21m09s | 2h26m |
| ProgramBench | $7.13 | $0.73 | 21m30s | 3h00m |
| Terminal-Bench 4.0 | $14.07 | $0.50 | 1h59m | 2h36m |
| Vibe Code Bench v1.1 | $4.83 | $1.04 | 23m43s | 1h01m |
| CyberBench v1.1 | $2.15 | $0.09 | 7m34s | 27m41s |
| SRE Bench | $21.33 | $0.53 | 1h50m | 2h23m |
Results available only for Gemini 3.7 Flash
- Vals RSI Index
- LegalBench
- MortgageTax
- TaxEval v2
- GPQA Diamond
- MMLU Pro
- MMMU Pro
- LiveCodeBench
- SkillsBench
- SWE-bench
Results available only for MiMo V2.6 Pro
- MysteryMechanism
- Public Benefits Bench v1.1