Compare / Claude Opus 4.5 vs GPT-5.2
Claude Opus 4.5 vs GPT-5.2
36 results from 6 sources that measured both models. Each row is in the source's own units; there is no overall winner.
| Benchmark | Metric | Claude Opus 4.5 | GPT-5.2 |
|---|---|---|---|
| Arena (formerly LMArena)reported | OverallArena rating | 1,465 | 1,425 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,187 | 1,162 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,459 | 1,442 |
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,472 | 1,444 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,469 | 1,391 |
| Arena (formerly LMArena)reported | Expert promptsArena rating | 1,503 | 1,475 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,483 | 1,427 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,474 | 1,437 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,467 | 1,411 |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Irrelevance detection% correct | 90.8% | 87.3% |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Memory% correct | 73.8% | 45.8% |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Multi-turn tasks% correct | 68.4% | 43.8% |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Overall accuracy% correct | 77.5% | 55.9% |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Relevance detection% correct | 68.8% | 75.0% |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Single-turn calls (curated)% correct | 89.7% | 81.8% |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Single-turn calls (user-contributed)% correct | 79.8% | 70.4% |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Web search% correct | 84.5% | 75.5% |
| OpenHands Indexreported | Average% resolved | 60.6% | 58.8% |
| OpenHands Indexreported | Front end (SWE-Bench Multimodal)% resolved | 41.2% | 30.9% |
| OpenHands Indexreported | Greenfield (Commit0)% resolved | 37.5% | 50.0% |
| OpenHands Indexreported | Information gathering (GAIA)% resolved | 69.1% | 65.5% |
| OpenHands Indexreported | Issue resolution (SWE-Bench)% resolved | 76.6% | 74.6% |
| OpenHands Indexreported | Testing (SWT-Bench)% resolved | 78.5% | 73.2% |
| tau2-benchreported | Consistency (pass^4)% of tasks | 70.0% | 72.0% |
| tau2-benchreported | Task success (pass^1)% of tasks | 84.0% | 83.0% |
| tau2-benchreported | Consistency (pass^4)% of tasks | 11.3% | 18.6% |
| tau2-benchreported | Task success (pass^1)% of tasks | 24.7% | 32.2% |
| tau2-benchreported | Consistency (pass^4)% of tasks | 51.8% | 51.8% |
| tau2-benchreported | Task success (pass^1)% of tasks | 79.6% | 81.6% |
| tau2-benchreported | Consistency (pass^4)% of tasks | 78.1% | 71.9% |
| tau2-benchreported | Task success (pass^1)% of tasks | 92.3% | 89.7% |
| UGI Leaderboardreported | Requested-length error% off the requested word count | 3.0% | 76.0% |
| UGI Leaderboardreported | Style adherencescore from 0 to 1 | 0.39 | 0.40 |
| UGI Leaderboardreported | Writing scorescore out of 100 | 70.4 | 49.4 |
| Vectara Hallucination Leaderboardreported | Answer rate% of documents | 98.7% | 100.0% |
| Vectara Hallucination Leaderboardreported | Hallucination rate% of summaries | 10.9% | 8.4% |
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.