Compare / GLM 4.6V vs GLM 5.2
GLM 4.6V vs GLM 5.2
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 | GLM 4.6V | GLM 5.2 |
|---|---|---|---|
| Arena (formerly LMArena)reported | OverallArena rating | 1,392 | 1,463 |
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,370 | 1,463 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,350 | 1,455 |
| Arena (formerly LMArena)reported | Expert promptsArena rating | 1,394 | 1,500 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,367 | 1,470 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,379 | 1,476 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,357 | 1,459 |
| Artificial Analysisreported | Artificial Analysis Intelligence Indexindex score | 11.2 | 33.7 |
| Artificial Analysisreported | GPQA Diamond% of questions | 71.9% | 89.5% |
| Artificial Analysisreported | Humanity's Last Exam% of questions | 9.6% | 41.1% |
| Artificial Analysisreported | IFBench% of instructions | 30.1% | 73.3% |
| Artificial Analysisreported | Long-context reasoning (AA-LCR)% of questions | 48.7% | 78.3% |
| Artificial Analysisreported | Terminal-Bench Hard% of tasks | 14.4% | 50.8% |
| Artificial Analysisreported | Τ²-bench telecom% of tasks | 31.6% | 99.1% |
| UGI Leaderboardreported | Requested-length error% off the requested word count | 35.0% | 12.0% |
| UGI Leaderboardreported | Style adherencescore from 0 to 1 | 0.39 | 0.37 |
| UGI Leaderboardreported | Writing scorescore out of 100 | 39.7 | 67.0 |
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.