Model comparison
Claude Sonnet 5 vs MiMo V2.6 Pro: Benchmark Comparison
Claude Sonnet 5 has the higher score on 6 of 18 shared benchmarks; MiMo V2.6 Pro leads on 11.
The largest observed score gap is 21.72 pts on Terminal-Bench 4.0 , where MiMo V2.6 Pro 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 Pro | Gap | Reported uncertainty |
|---|---|---|---|---|
| Vals Index | 51.77% ±1.09 | 55.20% ±1.18 | 3.42 pts | Reported ±1 SE ranges do not overlap |
| Harvey's Legal Agent Benchmark | 5.00% ±1.65 | 10.83% ±2.17 | 5.83 pts | Reported ±1 SE ranges do not overlap |
| Legal Research Bench | 41.83% ±3.43 | 47.12% ±3.47 | 5.29 pts | Reported ±1 SE ranges overlap |
| EMB | 66.32% ±3.01 | 62.86% ±3.14 | 3.46 pts | Reported ±1 SE ranges overlap |
| Finance Agent (v2) | 53.91% ±0.52 | 57.34% ±0.57 | 3.43 pts | Reported ±1 SE ranges do not overlap |
| Tax Agent Bench | 62.27% ±3.19 | 64.94% ±3.21 | 2.67 pts | Reported ±1 SE ranges overlap |
| MedCode | 47.54% ±2.27 | 44.97% ±2.10 | 2.57 pts | Reported ±1 SE ranges overlap |
| MedScribe | 76.05% ±3.05 | 88.31% ±1.94 | 12.25 pts | Reported ±1 SE ranges do not overlap |
| ProofBench v1.1 | 77.00% ±4.23 | 70.00% ±4.61 | 7.00 pts | Reported ±1 SE ranges overlap |
| Terminal-Bench Science | 2.86% ±2.01 | 2.86% ±2.01 | 0.00 pts | Reported ±1 SE ranges overlap |
| SAGE | 48.92% ±3.40 | 45.05% ±3.40 | 3.88 pts | Reported ±1 SE ranges overlap |
| Code Migration | 44.39% ±4.25 | 43.01% ±4.32 | 1.38 pts | Reported ±1 SE ranges overlap |
| IOI | 45.00% ±2.75 | 39.33% ±2.41 | 5.67 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 | 9.60% ±1.01 | 31.31% ±3.07 | 21.72 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench v1.1 | 81.33% ±3.05 | 85.22% ±3.39 | 3.90 pts | Reported ±1 SE ranges overlap |
| CyberBench v1.1 | 61.91% ±5.74 | 72.86% ±5.50 | 10.95 pts | Reported ±1 SE ranges overlap |
| Public Benefits Bench v1.1 | 66.03% ±1.23 | 68.94% ±1.20 | 2.91 pts | Reported ±1 SE ranges do not overlap |
Performance by category
| Category | Claude Sonnet 5 average | MiMo V2.6 Pro average |
|---|---|---|
| Legal | 23.41% | 28.97% |
| Finance | 60.83% | 61.71% |
| Healthcare | 61.80% | 66.64% |
| Math | 77.00% | 70.00% |
| Science | 2.86% | 2.86% |
| Education | 48.92% | 45.05% |
| Coding | 36.06% | 39.88% |
| Beta | 61.91% | 72.86% |
| Social Mobility | 66.03% | 68.94% |
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 Pro cost | Claude Sonnet 5 latency | MiMo V2.6 Pro latency |
|---|---|---|---|---|
| Vals Index | $13.72 | $0.41 | 54m16s | 1h07m |
| Harvey's Legal Agent Benchmark | $8.95 | $0.22 | 38m52s | 25m49s |
| Legal Research Bench | $2.72 | $0.18 | 25m46s | 30m16s |
| EMB | $10.29 | $0.33 | 49m46s | 55m14s |
| Finance Agent (v2) | $0.75 | $0.20 | 13m12s | 10m48s |
| Tax Agent Bench | $1.79 | $0.13 | 19m19s | 25m48s |
| MedCode | N/A | N/A | 2m15s | 2m57s |
| MedScribe | N/A | N/A | 4m12s | 2m53s |
| ProofBench v1.1 | $1.37 | $0.25 | 14m58s | 1h02m |
| Terminal-Bench Science | $18.70 | $0.68 | 2h55m | 3h40m |
| SAGE | N/A | N/A | 7m14s | 3m40s |
| Code Migration | $35.31 | $0.69 | 1h57m | 2h49m |
| IOI | $12.89 | $0.84 | 59m44s | 2h26m |
| ProgramBench | $24.36 | $0.73 | 1h31m | 3h00m |
| Terminal-Bench 4.0 | $26.33 | $0.50 | 1h45m | 2h42m |
| Vibe Code Bench v1.1 | $25.39 | $1.04 | 1h07m | 1h01m |
| CyberBench v1.1 | $1.89 | $0.09 | 16m57s | 27m41s |
| Public Benefits Bench v1.1 | $1.29 | $0.08 | 21m09s | 37m14s |
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 Pro
- MysteryMechanism
- SRE Bench