Claude Fable 5 vs GPT-6.1 Sol: Benchmark Comparison
Claude Fable 5 has the higher score on 7 of 12 shared benchmarks; GPT-6.1 Sol leads on 5.
The largest observed score gap is 37.14 pts on Terminal-Bench-Science , where GPT-6.1 Sol leads.
Reported ±1 standard-error ranges overlap on 4 of 11 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 Fable 5 | GPT-6.1 Sol | Gap | Reported uncertainty |
|---|---|---|---|---|
| Vals RSI Index | 24.32% | 26.62% | 2.30 pts | Uncertainty comparison unavailable |
| Legal Research Bench | 49.52% ±3.48 | 38.46% ±3.38 | 11.06 pts | Reported ±1 SE ranges do not overlap |
| Tax Agent Bench | 38.28% ±3.50 | 21.75% ±2.98 | 16.53 pts | Reported ±1 SE ranges do not overlap |
| MedCode | 56.07% ±2.20 | 48.84% ±2.12 | 7.23 pts | Reported ±1 SE ranges do not overlap |
| MedScribe | 88.52% ±1.95 | 86.45% ±1.90 | 2.07 pts | Reported ±1 SE ranges overlap |
| ProofBench v1.1 | 95.00% ±2.19 | 99.00% ±1.00 | 4.00 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench-Science | 15.71% ±4.38 | 52.86% ±6.01 | 37.14 pts | Reported ±1 SE ranges do not overlap |
| SAGE | 51.89% ±3.40 | 46.54% ±3.46 | 5.35 pts | Reported ±1 SE ranges overlap |
| ProgramBench | 2.00% ±0.99 | 3.00% ±1.21 | 1.00 pts | Reported ±1 SE ranges overlap |
| Terminal-Bench 4.0 | 41.41% ±1.34 | 55.05% ±1.82 | 13.64 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench v1.1 | 90.35% ±2.10 | 88.93% ±1.89 | 1.42 pts | Reported ±1 SE ranges overlap |
| Public Benefits Bench v1.1 | 70.43% ±1.19 | 59.34% ±1.28 | 11.10 pts | Reported ±1 SE ranges do not overlap |
Performance by category
| Category | Claude Fable 5 average | GPT-6.1 Sol average |
|---|---|---|
| Legal | 49.52% | 38.46% |
| Finance | 38.28% | 21.75% |
| Healthcare | 72.30% | 67.65% |
| Math | 95.00% | 99.00% |
| Science | 15.71% | 52.86% |
| Education | 51.89% | 46.54% |
| Coding | 44.59% | 48.99% |
| Social Mobility | 70.43% | 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 Fable 5 cost | GPT-6.1 Sol cost | Claude Fable 5 latency | GPT-6.1 Sol latency |
|---|---|---|---|---|
| Vals RSI Index | $491.22 | $98.21 | 90h00m | 90h00m |
| Legal Research Bench | $9.79 | $3.38 | 22m18s | 34m20s |
| Tax Agent Bench | $6.70 | $1.32 | 14m27s | 20m14s |
| MedCode | N/A | N/A | 91.44s | 115.34s |
| MedScribe | N/A | N/A | 119.47s | 2m50s |
| ProofBench v1.1 | $5.45 | $0.23 | 13m43s | 4m45s |
| Terminal-Bench-Science | $51.99 | $3.44 | 1h59m | 1h23m |
| SAGE | N/A | N/A | 116.92s | 50.22s |
| ProgramBench | $75.68 | $2.87 | 2h37m | 1h10m |
| Terminal-Bench 4.0 | $30.34 | $1.72 | 1h08m | 45m12s |
| Vibe Code Bench v1.1 | $41.71 | $6.23 | 1h02m | 1h15m |
| Public Benefits Bench v1.1 | $4.59 | $0.69 | 21m17s | 50m24s |
Results available only for Claude Fable 5
- LegalBench
- MortgageTax
- TaxEval v2
- GPQA Diamond
- MMLU Pro
- MMMU Pro
- LiveCodeBench
- SWE-bench
Results available only for GPT-6.1 Sol
- Vals Finance Index
- Vals Index
- Vals Legal Index
- Vals Science Index
- HLAB-Vals-1.1
- ClaimBench
- EMB v1.1
- Finance Agent v2.1
- BioMysteryBench
- MysteryMechanism
- Robo-bench
- Code Migration v1.1
- IOI
- SWE-bench Multimodal
- CyberBench v1.1
- SRE Bench