Compare / GPT-4o (2024-08-06) vs GPT-5.1
GPT-4o (2024-08-06) vs GPT-5.1
9 results from 3 sources that measured both models. Each row is in the source's own units; there is no overall winner.
| Benchmark | Metric | GPT-4o (2024-08-06) | GPT-5.1 |
|---|---|---|---|
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,302 | 1,448 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,327 | 1,428 |
| Arena (formerly LMArena)reported | Expert promptsArena rating | 1,312 | 1,482 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,319 | 1,450 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,336 | 1,456 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,341 | 1,437 |
| SimpleQA Verified (Epoch AI)reported | Correct answers% of questions | 26.0% | 48.0% |
| Vectara Hallucination Leaderboardreported | Answer rate% of documents | 93.8% | 100.0% |
| Vectara Hallucination Leaderboardreported | Hallucination rate% of summaries | 9.6% | 10.9% |
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.