Compare / Claude Fable 5.1 vs GPT-5.1
Claude Fable 5.1 vs GPT-5.1
20 results from 5 sources that measured both models. Each row is in the source's own units; there is no overall winner.
| Benchmark | Metric | Claude Fable 5.1 | GPT-5.1 |
|---|---|---|---|
| Arena (formerly LMArena)reported | OverallArena rating | 1,493 | 1,416 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,501 | 1,459 |
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,485 | 1,448 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,484 | 1,428 |
| Arena (formerly LMArena)reported | Expert promptsArena rating | 1,516 | 1,482 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,497 | 1,450 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,501 | 1,456 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,496 | 1,437 |
| Artificial Analysisreported | Artificial Analysis Coding Indexindex score | 81.6 | 49.4 |
| Artificial Analysisreported | Artificial Analysis Intelligence Indexindex score | 53.4 | 24.7 |
| Artificial Analysisreported | GPQA Diamond% of questions | 93.7% | 87.3% |
| Artificial Analysisreported | Humanity's Last Exam% of questions | 59.1% | 28.5% |
| Artificial Analysisreported | Long-context reasoning (AA-LCR)% of questions | 85.3% | 80.0% |
| Artificial Analysisreported | Terminal-Bench 2.1% of tasks | 91.4% | 52.4% |
| Artificial Analysisreported | Τ-bench banking% of tasks | 47.2% | 15.9% |
| SimpleQA Verified (Epoch AI)reported | Correct answers% of questions | 70.8% | 48.0% |
| UGI Leaderboardreported | Requested-length error% off the requested word count | 3.0% | 113.0% |
| UGI Leaderboardreported | Style adherencescore from 0 to 1 | 0.43 | 0.38 |
| UGI Leaderboardreported | Writing scorescore out of 100 | 63.9 | 55.8 |
| Vending-Bench 2reported | Money after a yearUS dollars | $5,421.56 | $1,473.43 |
Bold green marks the better value on that metric. Where a source reports ranges that overlap, the difference may not be meaningful; see the benchmark page for ranges.