Compare / o3 vs GLM 4.6
o3 vs GLM 4.6
18 results from 3 sources that measured both models. Each row is in the source's own units; there is no overall winner.
| Benchmark | Metric | o3 | GLM 4.6 |
|---|---|---|---|
| Arena (formerly LMArena)reported | OverallArena rating | 1,405 | 1,447 |
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,426 | 1,412 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,383 | 1,400 |
| Arena (formerly LMArena)reported | Expert promptsArena rating | 1,449 | 1,439 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,403 | 1,412 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,432 | 1,424 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,396 | 1,405 |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Irrelevance detection% correct | 86.1% | 85.0% |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Memory% correct | 51.8% | 55.7% |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Multi-turn tasks% correct | 62.2% | 68.0% |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Overall accuracy% correct | 63.0% | 72.4% |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Relevance detection% correct | 93.8% | 75.0% |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Single-turn calls (curated)% correct | 81.9% | 87.6% |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Single-turn calls (user-contributed)% correct | 73.2% | 80.9% |
| Berkeley Function Calling Leaderboard (BFCL) V4reported | Web search% correct | 77.0% | 77.5% |
| UGI Leaderboardreported | Requested-length error% off the requested word count | 18.0% | 13.0% |
| UGI Leaderboardreported | Style adherencescore from 0 to 1 | 0.30 | 0.38 |
| UGI Leaderboardreported | Writing scorescore out of 100 | 66.8 | 53.3 |
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.