Claude Opus 5.5 vs GPT-6 Sol: Benchmark Comparison
Claude Opus 5.5 has the higher score on 17 of 18 shared benchmarks; GPT-6 Sol leads on 1.
The largest observed score gap is 29.53 pts on Tax Agent Bench , where Claude Opus 5.5 leads.
Reported ±1 standard-error ranges overlap on 4 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 5.5 | GPT-6 Sol | Gap | Reported uncertainty |
|---|---|---|---|---|
| Vals RSI Index | 37.31% | 28.11% | 9.20 pts | Uncertainty comparison unavailable |
| Vals Science Index | 58.65% ±2.32 | 45.00% ±2.26 | 13.65 pts | Reported ±1 SE ranges do not overlap |
| Legal Research Bench | 50.48% ±3.48 | 28.85% ±3.15 | 21.64 pts | Reported ±1 SE ranges do not overlap |
| Tax Agent Bench | 44.54% ±3.58 | 15.00% ±2.58 | 29.53 pts | Reported ±1 SE ranges do not overlap |
| MedCode | 49.80% ±2.27 | 47.07% ±2.12 | 2.72 pts | Reported ±1 SE ranges overlap |
| MedScribe | 91.43% ±1.93 | 82.03% ±1.94 | 9.40 pts | Reported ±1 SE ranges do not overlap |
| ProofBench v1.1 | 100.00% ±0.00 | 83.00% ±3.77 | 17.00 pts | Reported ±1 SE ranges do not overlap |
| BioMysteryBench | 79.26% ±0.98 | 74.81% ±2.43 | 4.44 pts | Reported ±1 SE ranges do not overlap |
| MysteryMechanism | 49.55% ±3.36 | 30.18% ±3.09 | 19.37 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench-Science | 47.14% ±6.01 | 30.00% ±5.52 | 17.14 pts | Reported ±1 SE ranges do not overlap |
| IOI | 95.06% ±4.94 | 82.61% ±9.24 | 12.44 pts | Reported ±1 SE ranges overlap |
| ProgramBench | 18.50% ±2.75 | 2.00% ±0.99 | 16.50 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench 4.0 | 65.15% ±0.00 | 44.44% ±3.54 | 20.71 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench v1.1 | 90.29% ±1.53 | 87.82% ±2.53 | 2.47 pts | Reported ±1 SE ranges overlap |
| CyberBench v1.1 | 55.36% ±5.13 | 77.98% ±5.11 | 22.62 pts | Reported ±1 SE ranges do not overlap |
| SRE Bench | 33.59% ±2.92 | 30.53% ±2.85 | 3.05 pts | Reported ±1 SE ranges overlap |
| Financial Coaching Bench | 76.16% ±1.52 | 62.61% ±1.72 | 13.55 pts | Reported ±1 SE ranges do not overlap |
| Public Benefits Bench v1.1 | 70.64% ±1.19 | 56.63% ±1.29 | 14.00 pts | Reported ±1 SE ranges do not overlap |
Performance by category
| Category | Claude Opus 5.5 average | GPT-6 Sol average |
|---|---|---|
| Legal | 50.48% | 28.85% |
| Finance | 44.54% | 15.00% |
| Healthcare | 70.61% | 64.55% |
| Math | 100.00% | 83.00% |
| Science | 58.65% | 45.00% |
| Coding | 67.25% | 54.22% |
| Cyber | 44.47% | 54.25% |
| Social Mobility | 73.40% | 59.62% |
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 5.5 cost | GPT-6 Sol cost | Claude Opus 5.5 latency | GPT-6 Sol latency |
|---|---|---|---|---|
| Vals RSI Index | $411.86 | $665.11 | 90h00m | 90h00m |
| Vals Science Index | $8.71 | $2.31 | 59m51s | 31m16s |
| Legal Research Bench | $23.61 | $4.90 | 1h52m | 24m09s |
| Tax Agent Bench | $14.40 | $2.28 | 1h07m | 17m32s |
| MedCode | N/A | N/A | 4m07s | 62.71s |
| MedScribe | N/A | N/A | 7m22s | 81.14s |
| ProofBench v1.1 | $0.92 | $0.48 | 5m51s | 5m05s |
| BioMysteryBench | $2.95 | $0.66 | 12m19s | 5m24s |
| MysteryMechanism | $4.06 | $0.46 | 19m38s | 6m13s |
| Terminal-Bench-Science | $19.12 | $5.82 | 2h28m | 1h22m |
| IOI | $4.35 | $2.70 | 17m15s | 30m38s |
| ProgramBench | $42.66 | $11.67 | 2h24m | 46m11s |
| Terminal-Bench 4.0 | $12.04 | $5.79 | 1h04m | 34m56s |
| Vibe Code Bench v1.1 | $35.77 | $26.36 | 1h36m | 38m56s |
| CyberBench v1.1 | $2.85 | $1.34 | 18m11s | 12m14s |
| SRE Bench | $26.48 | $20.65 | 1h37m | 1h03m |
| Financial Coaching Bench | $1.63 | $0.23 | 10m52s | 6m51s |
| Public Benefits Bench v1.1 | $5.29 | $2.49 | 1h20m | 31m02s |
Results available only for Claude Opus 5.5
- Vals Index
- Vals Finance Index
- Vals Legal Index
- HLAB-Vals-1.1
- Immigration
- ClaimBench
- EMB v1.1
- Finance Agent v2.1
- Robo-bench
- SAGE
- Code Migration v1.1
- SWE-bench Multimodal
- Vibe Code Bench 1-100 v1.1
- CyberOffense
- Access to Justice Benchmark
- CUA-bench
- Time Horizon Index: KSP
Results available only for GPT-6 Sol
None.