Compare / Qwen3 235B A22B Instruct 2507 vs Qwen3 VL 235B A22B Thinking
Qwen3 235B A22B Instruct 2507 vs Qwen3 VL 235B A22B Thinking
16 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 235B A22B Instruct 2507 | Qwen3 VL 235B A22B Thinking |
|---|---|---|---|
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,432 | 1,392 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,378 | 1,339 |
| Arena (formerly LMArena)reported | Expert promptsArena rating | 1,448 | 1,439 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,414 | 1,383 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,422 | 1,395 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,396 | 1,365 |
| Artificial Analysisreported | Artificial Analysis Intelligence Indexindex score | 12.0 | 13.4 |
| Artificial Analysisreported | GPQA Diamond% of questions | 75.3% | 77.2% |
| Artificial Analysisreported | Humanity's Last Exam% of questions | 11.1% | 11.9% |
| Artificial Analysisreported | IFBench% of instructions | 46.1% | 56.5% |
| Artificial Analysisreported | Long-context reasoning (AA-LCR)% of questions | 33.9% | 63.7% |
| Artificial Analysisreported | Terminal-Bench Hard% of tasks | 15.2% | 11.4% |
| Artificial Analysisreported | Τ²-bench telecom% of tasks | 33.3% | 54.1% |
| UGI Leaderboardreported | Requested-length error% off the requested word count | 15.0% | 30.0% |
| UGI Leaderboardreported | Style adherencescore from 0 to 1 | 0.36 | 0.34 |
| UGI Leaderboardreported | Writing scorescore out of 100 | 49.0 | 32.3 |
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.