Model comparison
Claude Sonnet 5 vs MiMo V2.6 Flash: Benchmark Comparison
Claude Sonnet 5 has the higher score on 8 of 18 shared benchmarks; MiMo V2.6 Flash leads on 10.
The largest observed score gap is 14.65 pts on Terminal-Bench 4.0 , where MiMo V2.6 Flash leads.
Reported ±1 standard-error ranges overlap on 11 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 | Claude Sonnet 5 | MiMo V2.6 Flash | Gap | Reported uncertainty |
|---|---|---|---|---|
| Vals Index | 51.77% ±1.09 | 53.23% ±1.10 | 1.46 pts | Reported ±1 SE ranges overlap |
| Harvey's Legal Agent Benchmark | 5.00% ±1.65 | 11.25% ±2.29 | 6.25 pts | Reported ±1 SE ranges do not overlap |
| Legal Research Bench | 41.83% ±3.43 | 37.98% ±3.37 | 3.85 pts | Reported ±1 SE ranges overlap |
| EMB | 66.32% ±3.01 | 65.46% ±2.71 | 0.86 pts | Reported ±1 SE ranges overlap |
| Finance Agent (v2) | 53.91% ±0.52 | 56.28% ±0.46 | 2.37 pts | Reported ±1 SE ranges do not overlap |
| Tax Agent Bench | 62.27% ±3.19 | 59.90% ±3.13 | 2.37 pts | Reported ±1 SE ranges overlap |
| MedCode | 47.54% ±2.27 | 41.06% ±2.03 | 6.48 pts | Reported ±1 SE ranges do not overlap |
| MedScribe | 76.05% ±3.05 | 85.28% ±1.99 | 9.22 pts | Reported ±1 SE ranges do not overlap |
| ProofBench v1.1 | 77.00% ±4.23 | 63.00% ±4.85 | 14.00 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench Science | 2.86% ±2.01 | 4.29% ±2.44 | 1.43 pts | Reported ±1 SE ranges overlap |
| SAGE | 48.92% ±3.40 | 43.53% ±3.43 | 5.39 pts | Reported ±1 SE ranges overlap |
| Code Migration | 44.39% ±4.25 | 40.93% ±4.35 | 3.47 pts | Reported ±1 SE ranges overlap |
| IOI | 45.00% ±2.75 | 47.67% ±4.79 | 2.67 pts | Reported ±1 SE ranges overlap |
| ProgramBench | 0.00% ±0.00 | 0.50% ±0.50 | 0.50 pts | Reported ±1 SE ranges overlap |
| Terminal-Bench 4.0 | 9.60% ±1.01 | 24.24% ±1.51 | 14.65 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench v1.1 | 81.33% ±3.05 | 78.96% ±4.04 | 2.37 pts | Reported ±1 SE ranges overlap |
| CyberBench v1.1 | 61.91% ±5.74 | 75.36% ±5.42 | 13.45 pts | Reported ±1 SE ranges do not overlap |
| Public Benefits Bench v1.1 | 66.03% ±1.23 | 67.59% ±1.22 | 1.56 pts | Reported ±1 SE ranges overlap |
Performance by category
| Category | Claude Sonnet 5 average | MiMo V2.6 Flash average |
|---|---|---|
| Legal | 23.41% | 24.62% |
| Finance | 60.83% | 60.55% |
| Healthcare | 61.80% | 63.17% |
| Math | 77.00% | 63.00% |
| Science | 2.86% | 4.29% |
| Education | 48.92% | 43.53% |
| Coding | 36.06% | 38.46% |
| Beta | 61.91% | 75.36% |
| Social Mobility | 66.03% | 67.59% |
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 | Claude Sonnet 5 cost | MiMo V2.6 Flash cost | Claude Sonnet 5 latency | MiMo V2.6 Flash latency |
|---|---|---|---|---|
| Vals Index | $13.72 | $0.20 | 54m16s | 58m02s |
| Harvey's Legal Agent Benchmark | $8.95 | $0.09 | 38m52s | 17m12s |
| Legal Research Bench | $2.72 | $0.07 | 25m46s | 15m20s |
| EMB | $10.29 | $0.11 | 49m46s | 25m22s |
| Finance Agent (v2) | $0.75 | $0.07 | 13m12s | 7m02s |
| Tax Agent Bench | $1.79 | $0.05 | 19m19s | 17m19s |
| MedCode | N/A | N/A | 2m15s | 93.37s |
| MedScribe | N/A | N/A | 4m12s | 74.48s |
| ProofBench v1.1 | $1.37 | $0.12 | 14m58s | 1h03m |
| Terminal-Bench Science | $18.70 | $0.24 | 2h55m | 3h59m |
| SAGE | N/A | N/A | 7m14s | 2m16s |
| Code Migration | $35.31 | $0.49 | 1h57m | 3h12m |
| IOI | $12.89 | $0.22 | 59m44s | 1h49m |
| ProgramBench | $24.36 | $0.47 | 1h31m | 3h22m |
| Terminal-Bench 4.0 | $26.33 | $0.22 | 1h45m | 2h25m |
| Vibe Code Bench v1.1 | $25.39 | $0.56 | 1h07m | 50m13s |
| CyberBench v1.1 | $1.89 | $0.05 | 16m57s | 28m22s |
| Public Benefits Bench v1.1 | $1.29 | $0.03 | 21m09s | 20m56s |
Results available only for Claude Sonnet 5
- LegalBench
- MortgageTax
- TaxEval v2
- GPQA Diamond
- MMLU Pro
- MMMU Pro
- LiveCodeBench
- SkillsBench
- SWE-bench
- Vibe Code Bench 1-100
Results available only for MiMo V2.6 Flash
- BioMysteryBench
- MysteryMechanism
- SRE Bench