Compare / Qwen3 30B A3B Instruct 2507 vs o4 Mini
Qwen3 30B A3B Instruct 2507 vs o4 Mini
17 results from 3 sources that measured both models. Each row is in the source's own units; there is no overall winner.
| Benchmark | Metric | Qwen3 30B A3B Instruct 2507 | o4 Mini |
|---|---|---|---|
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,397 | 1,379 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,319 | 1,337 |
| Arena (formerly LMArena)reported | Expert promptsArena rating | 1,395 | 1,408 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,366 | 1,369 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,383 | 1,391 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,342 | 1,354 |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Irrelevance detection% correct | 79.9% | 87.2% |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Memory% correct | 17.6% | 35.3% |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Multi-turn tasks% correct | 30.0% | 41.8% |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Overall accuracy% correct | 41.4% | 53.2% |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Relevance detection% correct | 93.8% | 81.2% |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Single-turn calls (curated)% correct | 88.9% | 81.3% |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Single-turn calls (user-contributed)% correct | 78.4% | 70.8% |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Web search% correct | 22.5% | 75.5% |
| UGI Leaderboardreported | Requested-length error% off the requested word count | 13.0% | 19.0% |
| UGI Leaderboardreported | Style adherencescore from 0 to 1 | 0.31 | 0.33 |
| UGI Leaderboardreported | Writing scorescore out of 100 | 27.9 | 46.9 |
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.