GPT-5.6 Sol vs MiMo V2.6 Pro: Benchmark Comparison
GPT-5.6 Sol has the higher score on 10 of 15 shared benchmarks; MiMo V2.6 Pro leads on 5.
The largest observed score gap is 51.83 pts on IOI , where GPT-5.6 Sol leads.
Reported ±1 standard-error ranges overlap on 7 of 15 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 | GPT-5.6 Sol | MiMo V2.6 Pro | Gap | Reported uncertainty |
|---|---|---|---|---|
| Legal Research Bench | 48.08% ±3.47 | 47.12% ±3.47 | 0.96 pts | Reported ±1 SE ranges overlap |
| Tax Agent Bench | 31.09% ±3.34 | 32.39% ±3.36 | 1.30 pts | Reported ±1 SE ranges overlap |
| MedCode | 43.97% ±2.26 | 44.97% ±2.10 | 1.00 pts | Reported ±1 SE ranges overlap |
| MedScribe | 85.23% ±1.97 | 88.31% ±1.94 | 3.07 pts | Reported ±1 SE ranges overlap |
| ProofBench v1.1 | 83.00% ±3.77 | 70.00% ±4.61 | 13.00 pts | Reported ±1 SE ranges do not overlap |
| MysteryMechanism | 33.33% ±3.17 | 15.31% ±2.42 | 18.02 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench-Science | 20.00% ±4.82 | 2.86% ±2.01 | 17.14 pts | Reported ±1 SE ranges do not overlap |
| SAGE | 52.56% ±3.42 | 45.05% ±3.40 | 7.51 pts | Reported ±1 SE ranges do not overlap |
| IOI | 91.17% ±4.51 | 39.33% ±2.41 | 51.83 pts | Reported ±1 SE ranges do not overlap |
| ProgramBench | 1.50% ±0.86 | 0.50% ±0.50 | 1.00 pts | Reported ±1 SE ranges overlap |
| Terminal-Bench 4.0 | 37.88% ±0.00 | 31.31% ±3.07 | 6.57 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench v1.1 | 80.50% ±3.72 | 85.22% ±3.39 | 4.73 pts | Reported ±1 SE ranges overlap |
| CyberBench v1.1 | 76.31% ±5.18 | 72.86% ±5.50 | 3.45 pts | Reported ±1 SE ranges overlap |
| SRE Bench | 30.53% ±2.85 | 3.05% ±1.06 | 27.48 pts | Reported ±1 SE ranges do not overlap |
| Public Benefits Bench v1.1 | 66.51% ±1.23 | 68.94% ±1.20 | 2.44 pts | Reported ±1 SE ranges do not overlap |
Performance by category
| Category | GPT-5.6 Sol average | MiMo V2.6 Pro average |
|---|---|---|
| Legal | 48.08% | 47.12% |
| Finance | 31.09% | 32.39% |
| Healthcare | 64.60% | 66.64% |
| Math | 83.00% | 70.00% |
| Science | 26.67% | 9.09% |
| Education | 52.56% | 45.05% |
| Coding | 52.76% | 39.09% |
| Cyber | 53.42% | 37.95% |
| Social Mobility | 66.51% | 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 | GPT-5.6 Sol cost | MiMo V2.6 Pro cost | GPT-5.6 Sol latency | MiMo V2.6 Pro latency |
|---|---|---|---|---|
| Legal Research Bench | $21.61 | $0.18 | 1h17m | 30m16s |
| Tax Agent Bench | $7.51 | $0.13 | 44m41s | 25m48s |
| MedCode | N/A | N/A | 96.59s | 2m57s |
| MedScribe | N/A | N/A | 94.40s | 2m53s |
| ProofBench v1.1 | $1.44 | $0.25 | 8m40s | 1h02m |
| MysteryMechanism | $0.96 | $0.15 | 8m43s | 44m00s |
| Terminal-Bench-Science | $7.32 | $0.62 | 1h44m | 3h12m |
| SAGE | N/A | N/A | 75.39s | 3m40s |
| IOI | $7.98 | $0.84 | 1h02m | 2h26m |
| ProgramBench | $15.29 | $0.73 | 29m24s | 3h00m |
| Terminal-Bench 4.0 | $7.98 | $0.50 | 33m46s | 2h42m |
| Vibe Code Bench v1.1 | $33.40 | $1.04 | 33m53s | 1h01m |
| CyberBench v1.1 | $3.05 | $0.09 | 10m47s | 27m41s |
| SRE Bench | $42.52 | $0.53 | 1h28m | 2h23m |
| Public Benefits Bench v1.1 | $8.85 | $0.08 | 1h01m | 37m14s |
Results available only for GPT-5.6 Sol
- Vals RSI Index
- Vals Science Index
- LegalBench
- MortgageTax
- TaxEval v2
- BioMysteryBench
- GPQA Diamond
- MMLU Pro
- MMMU Pro
- LiveCodeBench
- SkillsBench
- SWE-bench
- Vibe Code Bench 1-100
- CUA-bench
- Time Horizon Index: KSP
Results available only for MiMo V2.6 Pro
- Vals Finance Index
- Vals Index
- Vals Legal Index
- HLAB-Vals-1.1
- EMB v1.1
- Finance Agent v2.1
- Code Migration v1.1
- SWE-bench Multimodal