Compare / Qwen3.6 Plus vs GPT-5
Qwen3.6 Plus vs GPT-5
21 results from 4 sources that measured both models. Each row is in the source's own units; there is no overall winner.
| Benchmark | Metric | Qwen3.6 Plus | GPT-5 |
|---|---|---|---|
| Arena (formerly LMArena)reported | OverallArena rating | 1,438 | 1,421 |
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,446 | 1,413 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,407 | 1,373 |
| Arena (formerly LMArena)reported | Expert promptsArena rating | 1,476 | 1,460 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,436 | 1,410 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,443 | 1,435 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,423 | 1,397 |
| Artificial Analysisreported | Artificial Analysis Coding Indexindex score | 54.5 | 37.8 |
| Artificial Analysisreported | Artificial Analysis Intelligence Indexindex score | 27.0 | 23.0 |
| Artificial Analysisreported | GPQA Diamond% of questions | 88.2% | 85.4% |
| Artificial Analysisreported | Humanity's Last Exam% of questions | 27.8% | 28.5% |
| Artificial Analysisreported | IFBench% of instructions | 75.2% | 73.1% |
| Artificial Analysisreported | Long-context reasoning (AA-LCR)% of questions | 78.3% | 78.2% |
| Artificial Analysisreported | Terminal-Bench 2.1% of tasks | 61.4% | 35.2% |
| Artificial Analysisreported | Terminal-Bench Hard% of tasks | 43.9% | 37.9% |
| Artificial Analysisreported | Τ-bench banking% of tasks | 20.8% | 22.1% |
| Artificial Analysisreported | Τ²-bench telecom% of tasks | 97.7% | 86.5% |
| SimpleQA Verified (Epoch AI)reported | Correct answers% of questions | 44.1% | 50.1% |
| UGI Leaderboardreported | Requested-length error% off the requested word count | 8.0% | 58.0% |
| UGI Leaderboardreported | Style adherencescore from 0 to 1 | 0.35 | 0.39 |
| UGI Leaderboardreported | Writing scorescore out of 100 | 58.2 | 53.5 |
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.