Compare / Qwen3 235B A22B Thinking 2507 vs GPT-5 Nano
Qwen3 235B A22B Thinking 2507 vs GPT-5 Nano
14 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 Thinking 2507 | GPT-5 Nano |
|---|---|---|---|
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,404 | 1,339 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,370 | 1,247 |
| Arena (formerly LMArena)reported | Expert promptsArena rating | 1,466 | 1,354 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,385 | 1,327 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,400 | 1,338 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,377 | 1,289 |
| Artificial Analysisreported | Artificial Analysis Intelligence Indexindex score | 12.7 | 13.0 |
| Artificial Analysisreported | GPQA Diamond% of questions | 79.0% | 67.6% |
| Artificial Analysisreported | Humanity's Last Exam% of questions | 15.9% | 9.5% |
| Artificial Analysisreported | IFBench% of instructions | 51.2% | 67.6% |
| Artificial Analysisreported | Long-context reasoning (AA-LCR)% of questions | 72.0% | 45.0% |
| Artificial Analysisreported | Terminal-Bench Hard% of tasks | 13.6% | 17.4% |
| Artificial Analysisreported | Τ²-bench telecom% of tasks | 53.2% | 36.5% |
| SimpleQA Verified (Epoch AI)reported | Correct answers% of questions | 40.4% | 11.7% |
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.