Compare / DeepSeek V3.2 vs Gemini 3.1 Pro Preview
DeepSeek V3.2 vs Gemini 3.1 Pro Preview
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 | DeepSeek V3.2 | Gemini 3.1 Pro Preview |
|---|---|---|---|
| Arena (formerly LMArena)reported | OverallArena rating | 1,437 | 1,472 |
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,423 | 1,477 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,400 | 1,480 |
| Arena (formerly LMArena)reported | Expert promptsArena rating | 1,453 | 1,509 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,420 | 1,479 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,425 | 1,487 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,410 | 1,480 |
| OpenHands Indexreported | Average% resolved | 45.7% | 60.6% |
| OpenHands Indexreported | Front end (SWE-Bench Multimodal)% resolved | 27.9% | 44.1% |
| OpenHands Indexreported | Greenfield (Commit0)% resolved | 25.0% | 25.0% |
| OpenHands Indexreported | Information gathering (GAIA)% resolved | 50.3% | 81.8% |
| OpenHands Indexreported | Issue resolution (SWE-Bench)% resolved | 71.6% | 76.8% |
| OpenHands Indexreported | Testing (SWT-Bench)% resolved | 53.6% | 75.1% |
| UGI Leaderboardreported | Requested-length error% off the requested word count | 26.0% | 19.0% |
| UGI Leaderboardreported | Style adherencescore from 0 to 1 | 0.36 | 0.34 |
| UGI Leaderboardreported | Writing scorescore out of 100 | 57.3 | 72.2 |
| Vectara Hallucination Leaderboardreported | Answer rate% of documents | 92.6% | 99.4% |
| Vectara Hallucination Leaderboardreported | Hallucination rate% of summaries | 6.3% | 10.4% |
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.