- SWE-bench Verified (internal scaffold) 73.4
- SWE-bench Multilingual 67.2
- SWE-bench Pro (Qwen-corrected tasks) 49.5
- Terminal-Bench 2.0 (Terminus-2, avg of 5) 51.5
- Claw-Eval Avg 68.7
- Claw-Eval Pass^3 50
- SkillsBench Avg5 28.7
- QwenClawBench 52.6
- NL2Repo 29.4
- QwenWebBench 1397
- TAU3-Bench (gpt-5.2 user model) 67.2
- VITA-Bench (claude-4-sonnet judge) 35.6
- DeepPlanning 25.9
- Tool Decathlon 26.9
- MCPMark 37
- MCP-Atlas (public set) 62.8
- WideSearch 60.1
- MMLU-Pro 85.2
- MMLU-Redux 93.3
- SuperGPQA 64.7
- C-Eval 90
- GPQA 86
- HLE 21.4
- LiveCodeBench v6 80.4
- HMMT Feb 25 90.7
- HMMT Nov 25 89.1
- HMMT Feb 26 83.6
- IMOAnswerBench 78.9
- AIME26 (full AIME 2026 I and II) 92.7
- MMMU 81.7
- MMMU-Pro 75.3
- Mathvista(mini) 86.4
- ZEROBench_sub 34.4
- RealWorldQA 85.3
- MMBench EN-DEV-v1.1 92.8
- SimpleVQA 58.9
- HallusionBench 69.8
- OmniDocBench1.5 89.9
- CharXiv(RQ) 78
- CC-OCR 81.9
- AI2D_TEST 92.7
- RefCOCO(avg) 92
- ODInW13 50.8
- EmbSpatialBench 84.3
- RefSpatialBench 64.3
- Best of VideoMME(w sub.) 86.6 / VideoMME(w/o sub.) 82.5 86.6
- VideoMMMU 83.7
- MLVU 86.2
- MVBench 74.6
- LVBench 71.4
Qwen 3.6 35B-A3B
Qwen 3.6 35B-A3B places in the top three on ODinW-13.
Qwen Released 14 Apr 2026 70
Benchmarks with a score
0
Ranked first, out of five or more models
1
In the top three
Qwen Apr 2026 Qwen 3.6 35B-A3B
Open weights
143.8
#100 of 271
Epoch Capabilities Index
- Terminal-Bench 2.0 24.6% #37 /43
- MMMU-Pro 75.3% #33 /72
- Toolathlon 26.9% #35 /50
Where it ranks
Its best score on each benchmark, and where that ranks among every model with a score there. We rank each model once, at its best setting.
Knowledge and language
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| MMLU-Pro | 85.2% | #23 of 151 | Model card | |
| HallusionBench | 69.8% | #4 of 23 | Model card | |
| SEA-HELM | 73.9 | #20 of 62 | Official leaderboard | |
| MMLU-Redux | 93.3% | #13 of 38 | Model card | |
| C-Eval | 90% | #17 of 38 | Model card | |
| SuperGPQA | 64.7% | #14 of 26 | Model card | |
| Humanity's Last Exam | 21.4% | #97 of 147 | Model card | |
| Extended NYT Connections | 41.6% | #68 of 97 | Official leaderboard | |
| Judgemark v4 | 32.6 | #32 of 40 | Official leaderboard | |
| SimpleVQA | 58.9% | #19 of 23 | Model card | |
| WideSearch | 60.1% | #17 of 17 | Model card |
Math and science
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| GPQA Diamond | 84.8% | #73 of 219 | Epoch AI | |
| OTIS Mock AIME 2024-2025 | 86.7% | #72 of 202 | Epoch AI | |
| HMMT 2025 | 90.7% | #21 of 49 | Model card | |
| DTBench | 56.5% | #107 of 178 | Official leaderboard | |
| AIME 2026 | 92.7% | #21 of 34 | Model card | |
| HMMT Nov 2025 | 89.1% | #16 of 23 | Model card | |
| CritPt | 0.29% | #112 of 160 | Epoch AI | |
| IMO-AnswerBench | 78.9% | #26 of 37 | Model card | |
| Surface Evolver Bench | 44.4% | #19 of 26 | Epoch AI | |
| HMMT 2026 | 83.6% | #14 of 18 | Model card | |
| FrontierMath Tiers 1-3 v2 | 20.4% | #76 of 89 | Epoch AI |
Coding
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| LiveCodeBench v6 | 80.4% | #28 of 76 | Model card | |
| SWE-bench Verified | 73.4% | #56 of 150 | Model card | |
| SWE-bench Multilingual | 67.2% | #48 of 70 | Model card | |
| WeirdML | 34.5% | #102 of 130 | Epoch AI | |
| SWE-Bench Pro | 49.5% | #67 of 77 | Model card | |
| NL2Repo-Bench | 29.4% | #25 of 28 | Model card | |
| WebDev Arena | 1,250 | #95 of 105 | Epoch AI | |
| SWE-rebench | 24.7% | #12 of 13 | Official leaderboard | |
| AndroidBench | 37% | #32 of 32 | Official leaderboard | |
| QwenWebBench | 1,397 | #4 of 4 | Model card |
Agents and tools
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Claw-Eval | 68.7% | #6 of 14 | Model card | |
| GAIA | 56.1% | #28 of 57 | Official leaderboard | |
| SnakeBench | 24 | #90 of 183 | Official leaderboard | |
| Mystery Game Puzzles | 22% | #42 of 80 | Epoch AI | |
| VITA-Bench | 35.6% | #7 of 12 | Model card | |
| tau3-bench | 67.2% | #8 of 12 | Model card | |
| Toolathlon | 26.9% | #35 of 50 | Official leaderboard | |
| MCP Atlas | 62.8% | #29 of 40 | Model card | |
| DeepPlanning | 25.9% | #5 of 6 | Model card | |
| Terminal-Bench 2.0 | 24.6% | #37 of 43 | Official leaderboard | |
| SkillsBench | 28.7% | #13 of 13 | Model card | |
| Claw-Eval Pass^3 | 50% | #11 of 11 | Model card | |
| MCPMark | 37% | #10 of 10 | Model card | |
| QwenClawBench | 52.6% | #6 of 6 | Model card |
Vision, audio and robotics
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| OCRBench v2 | 65.5% | #4 of 65 | Official leaderboard | |
| MMMU | 81.7% | #13 of 97 | Model card | |
| MathVista | 86.4% | #9 of 64 | Model card | |
| MMBench-EN v1.1 | 92.8% | #4 of 17 | Model card | |
| RealWorldQA | 85.3% | #9 of 30 | Model card | |
| RefCOCO / RefCOCO+ / RefCOCOg | 92% | #6 of 19 | Model card | |
| AI2D | 92.7% | #14 of 44 | Model card | |
| ODinW-13 | 50.8% | #3 of 9 | Model card | |
| Video-MME | 86.6% | #15 of 42 | Model card | |
| ZeroBench | 34.4% | #9 of 25 | Model card | |
| MMMU-Pro | 75.3% | #33 of 72 | Model card | |
| Video-MMMU | 83.7% | #14 of 30 | Model card | |
| MLVU | 86.2% | #8 of 16 | Model card | |
| EmbSpatialBench | 84.3% | #5 of 10 | Model card | |
| CC-OCR | 81.9% | #4 of 8 | Model card | |
| CharXiv | 78% | #39 of 67 | Model card | |
| LVBench | 71.4% | #16 of 25 | Model card | |
| MVBench | 74.6% | #10 of 15 | Model card | |
| RefSpatial-Bench | 64.3% | #6 of 8 | Model card | |
| ParseBench | 44.1% | #34 of 42 | Official leaderboard |
Other
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| LMCA | 29.7% | #80 of 139 | Epoch AI | |
| Conceptual Reasoning Index | 39.7 | #99 of 149 | Official leaderboard | |
| ACCoRD | 53.0 | #108 of 150 | Official leaderboard | |
| Political Science LLM Benchmark | 64.3% | #21 of 29 | Official leaderboard |
Qwen's models over time
Qwen's models in color, against every other model on the Epoch Capabilities Index, by release date.
Its launch post
The benchmarks the lab chose to headline, with the scores it gave.