Qwen3.6 35B A3B — Benchmarks
Benchmark scores for Qwen3.6 35B A3B aggregated from public leaderboards, with how it ranks among open models. See hardware requirements for what you need to run it.
Overall rank: #47 of 78 open modelscomposite 40.8/100 across 10 benchmarks in 3 categories · methodology
Coding
| Benchmark | Score | Open rank | All models |
|---|---|---|---|
| SciCode | 35.8% | #38 / 62 | #132 / 161 |
| WeirdML | 34.5% | #35 / 46 | #132 / 161 |
| Terminal-Bench | 23.0% | #13 / 16 | #50 / 57 |
| Surface Evolver Bench | 44.4% | #8 / 13 | #21 / 27 |
Math
| Benchmark | Score | Open rank | All models |
|---|---|---|---|
| FrontierMath | 20.3% | #15 / 15 | #73 / 102 |
| AIME 2024/2025 | 86.7% | #18 / 82 | #78 / 270 |
Reasoning
| Benchmark | Score | Open rank | All models |
|---|---|---|---|
| Chess Puzzles | 26.0% | #5 / 65 | #62 / 205 |
| Mystery Game Puzzles | 22.0% | #5 / 17 | #47 / 122 |
| DTBench | 73.9% | #30 / 67 | #130 / 210 |
| LMCA | 29.7% | #19 / 53 | #112 / 172 |
| GPQA Diamond | 84.9% | #19 / 94 | #84 / 289 |
| CritPt | 0.3% | #42 / 62 | #134 / 172 |
Scores aggregated from public benchmark sources (each linked from the benchmark pages). llmrun does not run these benchmarks.