Compare / Claude Sonnet 4.6 vs DeepSeek V4 Pro 0423
Claude Sonnet 4.6 vs DeepSeek V4 Pro 0423
18 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 Sonnet 4.6 | DeepSeek V4 Pro 0423 |
|---|---|---|---|
| Arena (formerly LMArena)reported | OverallArena rating | 1,461 | 1,455 |
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,482 | 1,460 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,449 | 1,446 |
| Arena (formerly LMArena)reported | Expert promptsArena rating | 1,506 | 1,480 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,475 | 1,452 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,472 | 1,458 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,455 | 1,451 |
| OpenHands Indexreported | Average% resolved | 44.5% | 40.7% |
| OpenHands Indexreported | Front end (SWE-Bench Multimodal)% resolved | 30.9% | 36.8% |
| OpenHands Indexreported | Greenfield (Commit0)% resolved | 50.0% | 12.5% |
| OpenHands Indexreported | Information gathering (GAIA)% resolved | 13.3% | 12.7% |
| OpenHands Indexreported | Issue resolution (SWE-Bench)% resolved | 74.4% | 73.2% |
| OpenHands Indexreported | Testing (SWT-Bench)% resolved | 54.0% | 68.1% |
| UGI Leaderboardreported | Requested-length error% off the requested word count | 2.0% | 34.0% |
| UGI Leaderboardreported | Style adherencescore from 0 to 1 | 0.41 | 0.36 |
| UGI Leaderboardreported | Writing scorescore out of 100 | 64.5 | 68.4 |
| Vectara Hallucination Leaderboardreported | Answer rate% of documents | 99.9% | 97.2% |
| Vectara Hallucination Leaderboardreported | Hallucination rate% of summaries | 10.6% | 8.6% |
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.