Alibaba
availableShows if the model has enough results for an index.Qwen3.6-35B-A3B
Qwen3.6-35B-A3B is a reasoning model from Alibaba. 41 benchmarks count toward its score, in 7 categories.
IndexOverall score out of 100.50.6 ±3.0
CoverageShare of the index weight with results.95%
SpeedOutput tokens per second.63/s
Input / 1MUS dollars per 1M input tokens.$0.15
Output / 1MUS dollars per 1M output tokens.$1
ContextMaximum tokens in one request.262K
EloLMArena rating and rank.N/A
The index is a score out of 100. The ± range shows how much it can change.
CapabilitiesScore per category, out of 100.
Out of 100Results
41 counted| BenchmarkThe test name. | CategoryThe capability that the test measures. | ResultThe score from the publisher. | IndexThis result as a score out of 100. | RunThe settings of the run. | DateDate of the result. | Published byThe source of the result. |
|---|---|---|---|---|---|---|
| τ²-Bench Tool-Agent-User Evaluation | Agentic | 95.3% | 67.2 | — | — | Victor Barres et al. |
| AI2D test split | Multimodal | 92.7% | — | — | — | Qwen |
| AIME 2026 | Math | 92.7% | 52.5 | — | — | Qwen |
| RefCOCO average | Multimodal | 92.0% | — | — | — | RefCOCO dataset authors |
| Harvard-MIT Mathematics Tournament February 2025 | Math | 90.7% | 49.6 | — | — | Qwen |
| C-Eval | Knowledge | 90.0% | — | — | — | C-Eval authors |
| OmniDocBench 1.5 | Multimodal | 89.9% | — | — | — | OpenAI |
| Harvard-MIT Mathematics Tournament November 2025 | Math | 89.1% | — | — | — | Qwen |
| Video-MME with subtitle | Multimodal | 86.6% | — | — | — | Qwen |
| MLVU mean average | Multimodal | 86.2% | — | — | — | Qwen |
| Graduate-Level Google-Proof Q&A | Knowledge | 86.0% | 57.6 | — | — | David Rein et al. |
| RealWorldQA | Multimodal | 85.3% | 56.9 | — | — | Qwen |
| Massive Multitask Language Understanding Professional | Knowledge | 85.2% | 54.8 | — | — | Yubo Wang et al. |
| GPQA diamond | Knowledge | 84.8% | 56.6 | none effort | — | Epoch AI |
| Artificial Analysis GPQA Diamond | Knowledge | 84.1% | 54.7 | — | — | Artificial Analysis |
| VideoMMMU | Multimodal | 83.7% | — | — | — | Qwen |
| Harvard-MIT Mathematics Tournament February 2026 | Math | 83.6% | 52.7 | — | — | Qwen |
| Video-MME without subtitle | Multimodal | 82.5% | — | — | — | Qwen |
| CC-OCR | Multimodal | 81.9% | — | — | — | Qwen |
| Massive Multi-discipline Multimodal Understanding | Multimodal | 81.7% | 50.9 | — | — | MMMU authors |
| LiveCodeBench: Holistic and Contamination Free Evaluation of Large Language Models for Code | Coding | 80.4% | 57.4 | — | — | Naman Jain et al. |
| MMAnswerBench | Math | 78.9% | — | — | — | Qwen |
| CharXiv Reasoning | Multimodal | 78.0% | 54.9 | — | — | CharXiv authors |
| Massive Multi-discipline Multimodal Understanding Pro | Multimodal | 75.3% | 49.9 | — | — | MMMU-Pro authors |
| Artificial Analysis MMMU-Pro | Multimodal | 75.0% | 58.7 | — | — | Artificial Analysis |
| Software Engineering Benchmark Verified | Coding | 73.4% | 56.4 | — | — | Carlos E. Jimenez et al. |
| Artificial Analysis Long Context Reasoning | Reasoning | 71.7% | 57.8 | — | — | Artificial Analysis |
| OTIS Mock AIME 2024-2025 | Math | 68.9% | 50.1 | none effort | — | Epoch AI |
| Claw-Eval | Agentic | 68.7% | 66.4 | — | — | Bowen Ye et al. |
| τ³-Bench Tool-Agent-User Evaluation | Agentic | 67.2% | 52.3 | — | — | Sierra Research |
| SuperGPQA: Scaling LLM Evaluation Across 285 Graduate Disciplines | Knowledge | 64.7% | 51.0 | — | — | Xiaoxuan Du et al. |
| Artificial Analysis IFBench | Instruction | 64.4% | 54.4 | — | — | Artificial Analysis |
| MCP Atlas | Agentic | 62.8% | 55.5 | — | — | OpenAI |
| WideResearch | Agentic | 60.1% | 44.3 | — | — | Qwen |
| SimpleVQA | Multimodal | 58.9% | 49.2 | — | — | Z.AI |
| QwenClawBench | Agentic | 52.6% | 52.0 | — | — | Qwen |
| ODINW13 | Multimodal | 50.8% | — | — | — | Qwen |
| SWE-bench Pro | Coding | 49.5% | 51.9 | — | — | Xiang Deng et al. |
| Gert Labs Composite Game Benchmark | Agentic | 42.6% | 50.4 | — | — | Gert Labs |
| Artificial Analysis Coding Index | Coding | 41.9% | 48.5 | — | — | Artificial Analysis |
| Artificial Analysis SciCode | Coding | 36.6% | 43.7 | — | — | Artificial Analysis |
| VITA-Bench | Agentic | 35.6% | 52.4 | — | — | Meituan LongCat Team |
| NL2Repo | Coding | 29.4% | 47.8 | — | — | MiniMax |
| Toolathlon | Agentic | 26.9% | 43.0 | — | — | OpenAI |
| DeepPlanning | Agentic | 25.9% | — | — | — | DeepPlanning authors |
| Artificial Analysis Humanity's Last Exam | Knowledge | 22.2% | 48.6 | — | — | Artificial Analysis |
| Mystery Game Puzzles | Reasoning | 22.0% | 57.3 | none effort | — | Epoch AI |
| Humanity's Last Exam | Knowledge | 21.4% | 46.9 | — | — | Center for AI Safety et al. |
| FrontierMath-Tiers-1-3-v2-Private | Math | 20.4% | 42.9 | none effort | — | Epoch AI |
| GDPval-AA normalized | Agentic | 19.0% | 50.2 | — | — | Artificial Analysis |
| Artificial Analysis Omniscience Accuracy | Knowledge | 18.8% | 37.1 | — | — | Artificial Analysis |
| Artificial Analysis Intelligence Index | Knowledge | 18.2% | 45.2 | — | — | Artificial Analysis |
| Artificial Analysis Agentic Index | Agentic | 15.0% | 49.4 | — | — | Artificial Analysis |
| Chess Puzzles | Reasoning | 4.0% | 28.7 | none effort | — | Epoch AI |
| Critical Physics Tasks | Reasoning | 0.3% | 39.0 | — | — | Artificial Analysis |
41 benchmarks count, from 42 of 55 results. A grey row does not count. Too few models took that benchmark.