Claude Sonnet 5.5 vs GPT-6.1 Sol: Benchmark Comparison
Claude Sonnet 5.5 has the higher score on 21 of 28 shared benchmarks; GPT-6.1 Sol leads on 6.
The largest observed score gap is 20.61 pts on SRE Bench , where GPT-6.1 Sol leads.
Reported ±1 standard-error ranges overlap on 11 of 28 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 5.5 | GPT-6.1 Sol | Gap | Reported uncertainty |
|---|---|---|---|---|
| Vals Index | 70.85% ±0.93 | 63.94% ±1.07 | 6.91 pts | Reported ±1 SE ranges do not overlap |
| Vals Finance Index | 64.53% ±1.48 | 55.81% ±1.50 | 8.72 pts | Reported ±1 SE ranges do not overlap |
| Vals Legal Index | 68.02% ±1.81 | 68.72% ±1.77 | 0.70 pts | Reported ±1 SE ranges overlap |
| Vals Science Index | 59.12% ±2.31 | 59.63% ±2.31 | 0.51 pts | Reported ±1 SE ranges overlap |
| HLAB-Vals-1.1 | 10.42% ±2.57 | 6.25% ±1.83 | 4.17 pts | Reported ±1 SE ranges overlap |
| Immigration | 68.96% ±1.54 | 67.81% ±1.53 | 1.16 pts | Reported ±1 SE ranges overlap |
| Legal Research Bench | 48.08% ±3.47 | 38.46% ±3.38 | 9.61 pts | Reported ±1 SE ranges do not overlap |
| ClaimBench | 100.00% ±0.00 | 100.00% ±0.00 | 0.00 pts | Reported ±1 SE ranges overlap |
| EMB v1.1 | 75.90% ±2.47 | 71.09% ±2.75 | 4.81 pts | Reported ±1 SE ranges overlap |
| Finance Agent v2.1 | 67.28% ±2.02 | 61.82% ±2.11 | 5.46 pts | Reported ±1 SE ranges do not overlap |
| Tax Agent Bench | 41.96% ±3.56 | 21.75% ±2.98 | 20.21 pts | Reported ±1 SE ranges do not overlap |
| MedCode | 52.92% ±2.12 | 48.84% ±2.12 | 4.08 pts | Reported ±1 SE ranges overlap |
| MedScribe | 91.10% ±1.96 | 86.45% ±1.90 | 4.65 pts | Reported ±1 SE ranges do not overlap |
| ProofBench v1.1 | 100.00% ±0.00 | 99.00% ±1.00 | 1.00 pts | Reported ±1 SE ranges overlap |
| BioMysteryBench | 81.11% ±0.64 | 79.63% ±0.74 | 1.48 pts | Reported ±1 SE ranges do not overlap |
| MysteryMechanism | 49.10% ±3.36 | 46.40% ±3.35 | 2.70 pts | Reported ±1 SE ranges overlap |
| Terminal-Bench-Science | 47.14% ±6.01 | 52.86% ±6.01 | 5.71 pts | Reported ±1 SE ranges overlap |
| SAGE | 88.03% ±1.62 | 90.49% ±1.53 | 2.46 pts | Reported ±1 SE ranges overlap |
| Code Migration v1.1 | 86.88% ±1.52 | 77.32% ±2.98 | 9.55 pts | Reported ±1 SE ranges do not overlap |
| IOI | 83.06% ±3.74 | 96.89% ±3.11 | 13.83 pts | Reported ±1 SE ranges do not overlap |
| ProgramBench | 6.50% ±1.75 | 3.00% ±1.21 | 3.50 pts | Reported ±1 SE ranges do not overlap |
| SWE-bench Multimodal | 58.33% ±2.25 | 43.54% ±2.27 | 14.79 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench 4.0 | 64.14% ±1.01 | 55.05% ±1.82 | 9.09 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench v1.1 | 92.39% ±1.26 | 88.93% ±1.89 | 3.46 pts | Reported ±1 SE ranges do not overlap |
| CyberBench v1.1 | 59.58% ±5.21 | 39.29% ±2.74 | 20.30 pts | Reported ±1 SE ranges do not overlap |
| SRE Bench | 30.15% ±2.84 | 50.76% ±3.10 | 20.61 pts | Reported ±1 SE ranges do not overlap |
| Access to Justice Benchmark | 74.94% ±0.70 | 65.84% ±0.76 | 9.11 pts | Reported ±1 SE ranges do not overlap |
| Public Benefits Bench v1.1 | 67.19% ±1.22 | 59.34% ±1.28 | 7.85 pts | Reported ±1 SE ranges do not overlap |
Performance by category
| Category | Claude Sonnet 5.5 average | GPT-6.1 Sol average |
|---|---|---|
| Legal | 42.49% | 37.51% |
| Finance | 71.28% | 63.66% |
| Healthcare | 72.01% | 67.65% |
| Math | 100.00% | 99.00% |
| Science | 59.12% | 59.63% |
| Education | 88.03% | 90.49% |
| Coding | 65.22% | 60.79% |
| Cyber | 44.87% | 45.02% |
| Social Mobility | 71.06% | 62.59% |
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 5.5 cost | GPT-6.1 Sol cost | Claude Sonnet 5.5 latency | GPT-6.1 Sol latency |
|---|---|---|---|---|
| Vals Index | $16.18 | $3.61 | 1h26m | 44m24s |
| Vals Finance Index | $7.40 | $1.36 | 43m31s | 19m16s |
| Vals Legal Index | $13.37 | $2.75 | 1h15m | 31m40s |
| Vals Science Index | $8.50 | $1.36 | 1h09m | 33m11s |
| HLAB-Vals-1.1 | $17.71 | $3.54 | 1h13m | 40m26s |
| Immigration | $1.16 | $0.23 | 11m03s | 7m50s |
| Legal Research Bench | $13.66 | $3.38 | 1h33m | 34m20s |
| ClaimBench | $0.045922 | $0.029239 | 4.05s | 11.26s |
| EMB v1.1 | $6.53 | $1.14 | 36m45s | 18m28s |
| Finance Agent v2.1 | $6.95 | $1.63 | 34m26s | 19m05s |
| Tax Agent Bench | $8.73 | $1.32 | 59m21s | 20m14s |
| MedCode | N/A | N/A | 3m59s | 115.34s |
| MedScribe | N/A | N/A | 5m07s | 2m50s |
| ProofBench v1.1 | $0.52 | $0.23 | 5m48s | 4m45s |
| BioMysteryBench | $2.95 | $0.37 | 18m06s | 8m51s |
| MysteryMechanism | $2.98 | $0.28 | 23m30s | 7m17s |
| Terminal-Bench-Science | $19.57 | $3.44 | 2h46m | 1h23m |
| SAGE | $0.16 | $0.08 | 81.12s | 50.43s |
| Code Migration v1.1 | $43.88 | $9.51 | 4h07m | 1h39m |
| IOI | $5.81 | $1.26 | 45m51s | 31m31s |
| ProgramBench | $22.62 | $2.87 | 2h07m | 1h10m |
| SWE-bench Multimodal | $3.26 | $0.35 | 24m13s | 6m56s |
| Terminal-Bench 4.0 | $10.75 | $1.72 | 1h22m | 45m12s |
| Vibe Code Bench v1.1 | $19.73 | $6.23 | 1h01m | 1h15m |
| CyberBench v1.1 | $2.91 | $0.29 | 33m53s | 3m42s |
| SRE Bench | $17.27 | $2.69 | 1h55m | 32m32s |
| Access to Justice Benchmark | $2.37 | $0.69 | 17m34s | 12m27s |
| Public Benefits Bench v1.1 | $3.76 | $0.69 | 1h15m | 50m24s |
Results available only for Claude Sonnet 5.5
None.
Results available only for GPT-6.1 Sol
- Vals RSI Index
- Robo-bench