Claude Fable 5.1 vs GPT-5.6 Sol
Claude Fable 5.1 
GPT-5.6 Sol Verdict
- Claude Fable 5.1 (xhigh) leads on quality: 70.3 vs 68.0.
- Claude Fable 5.1 (xhigh) is stronger in composite, knowledge, long context, reasoning.
- GPT-5.6 Sol (high) is stronger in coding, agents & tools, instruction following, multimodal.
- GPT-5.6 Sol (high) is 2.5× cheaper ($8.00 vs $20.00 per 1M blended).
| Metric | Claude Fable 5.1 (xhigh) | GPT-5.6 Sol (high) |
|---|---|---|
| BenchLeader Index | 70.3 | 68.0 |
| Coding score | 71.3 | 72.0 |
| Composite score | 95.0 | 82.5 |
| Knowledge score | 84.1 | 73.7 |
| Long context score | 68.1 | 67.4 |
| Reasoning score | 76.4 | 62.9 |
| Agents & tools score | – | 87.6 |
| Instruction following score | – | 67.6 |
| Multimodal score | – | 66.5 |
| Blended price $/M | $20.00 | $8.00 |
| Output speed | 58 tok/s | 59 tok/s |
| Time to first answer | 132.0 s | 31.4 s |
| Context window | 1M | 1M |
| Humanity's Last Exam | 46.5% | – |
| SciCode | 60.1% | 56.9% |
| WeirdML | – | 88.8% |
| AA Intelligence Index | 53.2 | 42.5 |
| IFBench | – | 69.2% |
| AA-LCR | 83.0% | 81.7% |
| MMMU-Pro | – | 81.8% |
| AA-Omniscience | 42.4 | 20.4 |
| Terminal-Bench Hard | – | 62.1% |
| GPQA Diamond (AA) | 93.4% | 92.8% |
| Humanity's Last Exam (AA) | 58.7% | 46.0% |
| SciCode (AA) | 60.9% | 57.8% |
| τ²-Bench Telecom (AA) | – | 83.3% |
| EnigmaEval | – | 37.1% |
| ARC-AGI-1 | 96.5% | 97.0% |
| ARC-AGI-2 | 90.0% | 85.4% |
| ARC-AGI-3 | – | 2.2% |
Data as of 2026-09-09. Best configuration of each model; every score links to its source on the model pages.