Compare / Claude Opus 4.8 vs GPT-5.2
Claude Opus 4.8 vs GPT-5.2
20 results from 4 sources that measured both models. Each row is in the source's own units; there is no overall winner.
| Benchmark | Metric | Claude Opus 4.8 | GPT-5.2 |
|---|---|---|---|
| Arena (formerly LMArena)reported | OverallArena rating | 1,486 | 1,425 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,195 | 1,162 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,469 | 1,442 |
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,490 | 1,444 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,467 | 1,391 |
| Arena (formerly LMArena)reported | Expert promptsArena rating | 1,525 | 1,475 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,491 | 1,427 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,480 | 1,437 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,474 | 1,411 |
| OpenHands Indexreported | Average% resolved | 71.9% | 58.8% |
| OpenHands Indexreported | Front end (SWE-Bench Multimodal)% resolved | 50.0% | 30.9% |
| OpenHands Indexreported | Greenfield (Commit0)% resolved | 62.5% | 50.0% |
| OpenHands Indexreported | Information gathering (GAIA)% resolved | 78.8% | 65.5% |
| OpenHands Indexreported | Issue resolution (SWE-Bench)% resolved | 83.8% | 74.6% |
| OpenHands Indexreported | Testing (SWT-Bench)% resolved | 84.3% | 73.2% |
| tau2-benchreported | Consistency (pass^4)% of tasks | 22.7% | 18.6% |
| tau2-benchreported | Task success (pass^1)% of tasks | 39.7% | 32.2% |
| UGI Leaderboardreported | Requested-length error% off the requested word count | 6.0% | 76.0% |
| UGI Leaderboardreported | Style adherencescore from 0 to 1 | 0.40 | 0.40 |
| UGI Leaderboardreported | Writing scorescore out of 100 | 65.9 | 49.4 |
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.