Compare / Gemma 3 27B vs Gemma 4 31B
Gemma 3 27B vs Gemma 4 31B
12 results from 3 sources that measured both models. Each row is in the source's own units; there is no overall winner.
| Benchmark | Metric | Gemma 3 27B | Gemma 4 31B |
|---|---|---|---|
| Arena (formerly LMArena)reported | OverallArena rating | 1,380 | 1,450 |
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,377 | 1,449 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,348 | 1,419 |
| Arena (formerly LMArena)reported | Expert promptsArena rating | 1,337 | 1,499 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,343 | 1,453 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,365 | 1,453 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,354 | 1,435 |
| UGI Leaderboardreported | Requested-length error% off the requested word count | 8.0% | 5.0% |
| UGI Leaderboardreported | Style adherencescore from 0 to 1 | 0.29 | 0.36 |
| UGI Leaderboardreported | Writing scorescore out of 100 | 45.0 | 43.1 |
| Vectara Hallucination Leaderboardreported | Answer rate% of documents | 98.8% | 100.0% |
| Vectara Hallucination Leaderboardreported | Hallucination rate% of summaries | 7.4% | 7.4% |
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.