Claude Opus 5 vs GPT-6.1 Sol: Benchmark Comparison
Claude Opus 5 has the higher score on 8 of 18 shared benchmarks; GPT-6.1 Sol leads on 8.
The largest observed score gap is 38.55 pts on SRE Bench , where GPT-6.1 Sol leads.
Reported ±1 standard-error ranges overlap on 7 of 17 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 Opus 5 | GPT-6.1 Sol | Gap | Reported uncertainty |
|---|---|---|---|---|
| Vals RSI Index | 33.02% | 26.62% | 6.40 pts | Uncertainty comparison unavailable |
| Vals Science Index | 47.93% ±2.20 | 59.63% ±2.31 | 11.70 pts | Reported ±1 SE ranges do not overlap |
| Legal Research Bench | 55.29% ±3.46 | 38.46% ±3.38 | 16.83 pts | Reported ±1 SE ranges do not overlap |
| Tax Agent Bench | 45.55% ±3.59 | 21.75% ±2.98 | 23.79 pts | Reported ±1 SE ranges do not overlap |
| MedCode | 63.57% ±1.99 | 48.84% ±2.12 | 14.73 pts | Reported ±1 SE ranges do not overlap |
| MedScribe | 90.98% ±1.92 | 86.45% ±1.90 | 4.53 pts | Reported ±1 SE ranges do not overlap |
| ProofBench v1.1 | 99.00% ±1.00 | 99.00% ±1.00 | 0.00 pts | Reported ±1 SE ranges overlap |
| BioMysteryBench | 79.26% ±2.06 | 79.63% ±0.74 | 0.37 pts | Reported ±1 SE ranges overlap |
| MysteryMechanism | 37.39% ±3.25 | 46.40% ±3.35 | 9.01 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench-Science | 27.14% ±5.35 | 52.86% ±6.01 | 25.71 pts | Reported ±1 SE ranges do not overlap |
| SAGE | 49.43% ±3.29 | 46.54% ±3.46 | 2.89 pts | Reported ±1 SE ranges overlap |
| IOI | 84.33% ±9.96 | 96.89% ±3.11 | 12.56 pts | Reported ±1 SE ranges overlap |
| ProgramBench | 3.00% ±1.21 | 3.00% ±1.21 | 0.00 pts | Reported ±1 SE ranges overlap |
| Terminal-Bench 4.0 | 53.53% ±1.34 | 55.05% ±1.82 | 1.52 pts | Reported ±1 SE ranges overlap |
| Vibe Code Bench v1.1 | 88.40% ±3.00 | 88.93% ±1.89 | 0.53 pts | Reported ±1 SE ranges overlap |
| CyberBench v1.1 | 65.36% ±5.55 | 39.29% ±2.74 | 26.07 pts | Reported ±1 SE ranges do not overlap |
| SRE Bench | 12.21% ±2.03 | 50.76% ±3.10 | 38.55 pts | Reported ±1 SE ranges do not overlap |
| Public Benefits Bench v1.1 | 76.93% ±1.10 | 59.34% ±1.28 | 17.59 pts | Reported ±1 SE ranges do not overlap |
Performance by category
| Category | Claude Opus 5 average | GPT-6.1 Sol average |
|---|---|---|
| Legal | 55.29% | 38.46% |
| Finance | 45.55% | 21.75% |
| Healthcare | 77.28% | 67.65% |
| Math | 99.00% | 99.00% |
| Science | 47.93% | 59.63% |
| Education | 49.43% | 46.54% |
| Coding | 57.32% | 60.97% |
| Cyber | 38.79% | 45.02% |
| Social Mobility | 76.93% | 59.34% |
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 Opus 5 cost | GPT-6.1 Sol cost | Claude Opus 5 latency | GPT-6.1 Sol latency |
|---|---|---|---|---|
| Vals RSI Index | $418.69 | $98.21 | 90h00m | 90h00m |
| Vals Science Index | $13.04 | $1.36 | 1h02m | 33m11s |
| Legal Research Bench | $6.76 | $3.38 | 31m01s | 34m20s |
| Tax Agent Bench | $5.14 | $1.32 | 16m48s | 20m14s |
| MedCode | N/A | N/A | 24.50s | 115.34s |
| MedScribe | N/A | N/A | 76.56s | 2m50s |
| ProofBench v1.1 | $2.11 | $0.23 | 9m51s | 4m45s |
| BioMysteryBench | $3.29 | $0.37 | 36m36s | 8m51s |
| MysteryMechanism | $3.28 | $0.28 | 16m53s | 7m17s |
| Terminal-Bench-Science | $32.54 | $3.44 | 2h12m | 1h23m |
| SAGE | N/A | N/A | 56.17s | 50.22s |
| IOI | $16.48 | $1.26 | 56m42s | 31m31s |
| ProgramBench | $60.29 | $2.87 | 2h55m | 1h10m |
| Terminal-Bench 4.0 | $18.60 | $1.72 | 1h07m | 45m12s |
| Vibe Code Bench v1.1 | $33.88 | $6.23 | 1h28m | 1h15m |
| CyberBench v1.1 | $2.87 | $0.29 | 19m25s | 3m42s |
| SRE Bench | $23.63 | $2.69 | 1h21m | 32m32s |
| Public Benefits Bench v1.1 | $3.95 | $0.69 | 30m09s | 50m24s |
Results available only for Claude Opus 5
- LegalBench
- MortgageTax
- TaxEval v2
- 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 GPT-6.1 Sol
- Vals Finance Index
- Vals Index
- Vals Legal Index
- HLAB-Vals-1.1
- ClaimBench
- EMB v1.1
- Finance Agent v2.1
- Robo-bench
- Code Migration v1.1
- SWE-bench Multimodal