Compare / GLM 4.6V vs GLM 4.7
GLM 4.6V vs GLM 4.7
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 4.7 |
|---|---|---|---|
| Arena (formerly LMArena)reported | OverallArena rating | 1,392 | 1,444 |
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,370 | 1,442 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,350 | 1,404 |
| Arena (formerly LMArena)reported | Expert promptsArena rating | 1,394 | 1,449 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,367 | 1,425 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,379 | 1,441 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,357 | 1,416 |
| Artificial Analysisreported | Artificial Analysis Intelligence Indexindex score | 11.2 | 22.2 |
| Artificial Analysisreported | GPQA Diamond% of questions | 71.9% | 85.9% |
| Artificial Analysisreported | Humanity's Last Exam% of questions | 9.6% | 27.4% |
| Artificial Analysisreported | IFBench% of instructions | 30.1% | 67.9% |
| Artificial Analysisreported | Long-context reasoning (AA-LCR)% of questions | 48.7% | 71.0% |
| Artificial Analysisreported | Terminal-Bench Hard% of tasks | 14.4% | 31.8% |
| Artificial Analysisreported | Τ²-bench telecom% of tasks | 31.6% | 95.9% |
| UGI Leaderboardreported | Requested-length error% off the requested word count | 35.0% | 28.0% |
| UGI Leaderboardreported | Style adherencescore from 0 to 1 | 0.39 | 0.38 |
| UGI Leaderboardreported | Writing scorescore out of 100 | 39.7 | 47.4 |
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.