- SWE-bench Verified (internal scaffold) 77.2
- SWE-bench Pro (Qwen-corrected tasks) 53.5
- SWE-bench Multilingual 71.3
- Terminal-Bench 2.0 (Terminus-2, avg of 5) 59.3
- SkillsBench Avg5 (OpenCode, 78 tasks) 48.2
- QwenWebBench 1487
- NL2Repo 36.2
- Claw-Eval Avg 72.4
- Claw-Eval Pass^3 60.6
- QwenClawBench 53.4
- MMLU-Pro 86.2
- MMLU-Redux 93.5
- SuperGPQA 66
- C-Eval 91.4
- GPQA Diamond 87.8
- HLE 24
- LiveCodeBench v6 83.9
- HMMT Feb 25 93.8
- HMMT Nov 25 90.7
- HMMT Feb 26 84.3
- IMOAnswerBench 80.8
- AIME26 (full AIME 2026 I and II) 94.1
- MMMU 82.9
- MMMU-Pro 75.8
- MathVista mini 87.4
- DynaMath 85.6
- VlmsAreBlind 97
- RealWorldQA 84.1
- MMStar 81.4
- MMBench EN-DEV-v1.1 92.3
- SimpleVQA 56.1
- CharXiv RQ 78.4
- CC-OCR 81.2
- OCRBench 89.4
- ERQA 62.5
- CountBench 97.8
- RefCOCO avg 92.5
- EmbSpatialBench 84.6
- RefSpatialBench 70
- VideoMME(w sub.) 87.7
- VideoMMMU 84.4
- MLVU 86.6
- MVBench 75.5
- V* 94.7
- AndroidWorld 70.3
Qwen3.6 27B
Qwen3.6 27B ranks first on 2 of the 71 benchmarks it has scores on.
Qwen Released 22 Apr 2026 73
Benchmarks with a score
2
Ranked first, out of five or more models
6
In the top three
Qwen Apr 2026 Qwen3.6 27B
Open weights
146.3
#81 of 271
Epoch Capabilities Index
- MMMU-Pro 75.8% #31 /72
- NL2Repo-Bench 36.2% #24 /28
- SimpleVQA 56.1% #20 /23
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 | 86.2% | #16 of 151 | Model card | |
| SEA-HELM | 74.6 | #16 of 62 | Official leaderboard | |
| MMLU-Redux | 93.5% | #11 of 38 | Model card | |
| C-Eval | 91.4% | #12 of 38 | Model card | |
| SuperGPQA | 66% | #11 of 26 | Model card | |
| Judgemark v4 | 53.4 | #24 of 40 | Official leaderboard | |
| Humanity's Last Exam | 24% | #89 of 147 | Model card | |
| SimpleVQA | 56.1% | #20 of 23 | Model card | |
| EQ-Bench 4 | 1,015 | #31 of 32 | Official leaderboard |
Math and science
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| OTIS Mock AIME 2024-2025 | 91.1% | #58 of 202 | Epoch AI | |
| GPQA Diamond | 85.9% | #68 of 219 | Epoch AI | |
| HMMT 2025 | 93.8% | #16 of 49 | Model card | |
| AIME 2026 | 94.1% | #18 of 34 | Model card | |
| DTBench | 63.5% | #95 of 178 | Official leaderboard | |
| HMMT Nov 2025 | 90.7% | #13 of 23 | Model card | |
| CritPt | 0.86% | #99 of 160 | Epoch AI | |
| IMO-AnswerBench | 80.8% | #24 of 37 | Model card | |
| FrontierMath Tiers 1-3 v2 | 35.1% | #64 of 89 | Epoch AI | |
| HMMT 2026 | 84.3% | #13 of 18 | Model card | |
| ProofBench | 8.00% | #48 of 59 | Official leaderboard |
Coding
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| SWE-bench Verified | 77.2% | #34 of 150 | Model card | |
| LiveCodeBench v6 | 83.9% | #22 of 76 | Model card | |
| SWE-bench Multilingual | 71.3% | #41 of 70 | Model card | |
| SWE-Bench Pro | 53.5% | #55 of 77 | Model card | |
| QwenWebBench | 1,487 | #3 of 4 | Model card | |
| Vibe Code Bench | 11.9% | #95 of 115 | Official leaderboard | |
| SWE-rebench | 31.2% | #11 of 13 | Official leaderboard | |
| NL2Repo-Bench | 36.2% | #24 of 28 | Model card | |
| AndroidBench | 45.1% | #29 of 32 | Official leaderboard |
Agents and tools
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Claw-Eval | 72.4% | #4 of 14 | Model card | |
| Terminal-Bench 2.0 | 59.3% | #21 of 61 | Model card | |
| LLM Chess | 270 | #50 of 124 | Official leaderboard | |
| AndroidWorld | 70.3% | #8 of 18 | Model card | |
| Claw-Eval Pass^3 | 60.6% | #6 of 11 | Model card | |
| SkillsBench | 48.2% | #9 of 13 | Model card | |
| SnakeBench | 18 | #143 of 183 | Official leaderboard | |
| QwenClawBench | 53.4% | #5 of 6 | Model card | |
| LiveClawBench | 54.5% | #15 of 17 | Official leaderboard | |
| Mystery Game Puzzles | 7.00% | #74 of 80 | Epoch AI | |
| CivBench | 1,391 | #18 of 18 | Official leaderboard |
Professional work
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| MortgageTax | 68.3% | #18 of 86 | Official leaderboard | |
| CorpFin v2 | 62.3% | #49 of 119 | Official leaderboard | |
| TaxEval v2 | 71.3% | #74 of 129 | Official leaderboard | |
| CaseLaw v2 | 53.2% | #41 of 53 | Official leaderboard |
Vision, audio and robotics
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| OCRBench v2 | 68.3% | #1 of 65 | Official leaderboard | |
| CountBench | 97.8% | #1 of 20 | Model card | |
| MMMU | 82.9% | #10 of 97 | Model card | |
| MathVista | 87.4% | #7 of 64 | Model card | |
| RefCOCO / RefCOCO+ / RefCOCOg | 92.5% | #3 of 19 | Model card | |
| MMStar | 81.4% | #6 of 30 | Model card | |
| EmbSpatialBench | 84.6% | #2 of 10 | Model card | |
| DynaMath | 85.6% | #4 of 17 | Model card | |
| Video-MME | 87.7% | #10 of 42 | Model card | |
| OCRBench | 89.4% | #7 of 28 | Model card | |
| VLMs Are Blind | 97% | #2 of 8 | Model card | |
| MMBench-EN v1.1 | 92.3% | #5 of 17 | Model card | |
| RefSpatial-Bench | 70% | #3 of 8 | Model card | |
| RealWorldQA | 84.1% | #12 of 30 | Model card | |
| Video-MMMU | 84.4% | #12 of 30 | Model card | |
| MVBench | 75.5% | #6 of 15 | Model card | |
| V* | 94.7% | #5 of 12 | Model card | |
| MMMU-Pro | 75.8% | #31 of 72 | Model card | |
| MLVU | 86.6% | #7 of 16 | Model card | |
| CharXiv | 78.4% | #38 of 67 | Model card | |
| ERQA | 62.5% | #15 of 26 | Model card | |
| CC-OCR | 81.2% | #7 of 8 | Model card |
Safety and security
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Human Pathogen Capabilities Test v2 (HPCT v2) | 47.3% | #48 of 71 | Official leaderboard |
Other
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Political Science LLM Benchmark | 66.6% | #11 of 29 | Official leaderboard | |
| SAGE (Vals) | 45.6% | #39 of 81 | Official leaderboard | |
| LMCA | 34.5% | #72 of 139 | Epoch AI | |
| Conceptual Reasoning Index | 45.2 | #86 of 149 | Official leaderboard | |
| ACCoRD | 58.9 | #91 of 150 | Official leaderboard | |
| Decision Index | 54.7 | #3 of 3 | 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.