Compare / Qwen3 VL 235B A22B Thinking vs DeepSeek V3.1 Terminus
Qwen3 VL 235B A22B Thinking vs DeepSeek V3.1 Terminus
15 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 VL 235B A22B Thinking | DeepSeek V3.1 Terminus |
|---|---|---|---|
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,392 | 1,423 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,339 | 1,405 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,383 | 1,420 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,395 | 1,417 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,365 | 1,411 |
| Artificial Analysisreported | Artificial Analysis Intelligence Indexindex score | 13.4 | 14.8 |
| Artificial Analysisreported | GPQA Diamond% of questions | 77.2% | 79.2% |
| Artificial Analysisreported | Humanity's Last Exam% of questions | 11.9% | 16.4% |
| Artificial Analysisreported | IFBench% of instructions | 56.5% | 57.0% |
| Artificial Analysisreported | Long-context reasoning (AA-LCR)% of questions | 63.7% | 69.3% |
| Artificial Analysisreported | Terminal-Bench Hard% of tasks | 11.4% | 31.8% |
| Artificial Analysisreported | Τ²-bench telecom% of tasks | 54.1% | 37.1% |
| UGI Leaderboardreported | Requested-length error% off the requested word count | 30.0% | 32.0% |
| UGI Leaderboardreported | Style adherencescore from 0 to 1 | 0.34 | 0.37 |
| UGI Leaderboardreported | Writing scorescore out of 100 | 32.3 | 54.0 |
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.