| visual-grounding-on-refcoco-refcoco-refcocog-overall | Qwen3.6-27B | #1 | Accuracy: 92.5 |
| visual-grounding-on-refcoco-refcoco-refcocog-overall | Qwen3.6-35B-A3B | #4 | Accuracy: 92.0 |
| code-generation-on-terminal-bench-2-0-terminus-2 | Qwen3.6-27B (Terminus-2) | #2 | Accuracy: 59.3 |
| code-generation-on-terminal-bench-2-0-terminus-2 | Qwen3.6-35B-A3B (Terminus-2) | #5 | Accuracy: 51.5 |
| video-question-answering-on-mlvu | Qwen3.6-27B | #3 | Accuracy: 86.6 |
| video-question-answering-on-mlvu | Qwen3.6-35B-A3B | #4 | Accuracy: 86.2 |
| video-question-answering-on-video-mme-with-subtitles | Qwen3.6-27B | #3 | Accuracy: 87.7 |
| video-question-answering-on-video-mme-with-subtitles | Qwen3.6-35B-A3B | #9 | Accuracy: 86.6 |
| vision-language-understanding-on-mmbench-v1-1 | Qwen3.6-35B-A3B | #3 | Accuracy: 92.8 |
| vision-language-understanding-on-mmbench-v1-1 | Qwen3.6-27B | #5 | Accuracy: 92.3 |
| agents-on-claw-eval | Qwen3.6-27B | #4 | Accuracy: 72.4 |
| agents-on-claw-eval | Qwen3.6-35B-A3B | #6 | Accuracy: 68.7 |
| multiple-choice-question-answering-mcqa-on-supergpqa | Qwen3.6-27B | #4 | Accuracy: 66.0 |
| multiple-choice-question-answering-mcqa-on-supergpqa | Qwen3.6-35B-A3B | #6 | Accuracy: 64.7 |
| object-counting-on-countbench | Qwen3.6-27B | #4 | Accuracy: 97.8 |
| object-counting-on-countbench | Qwen3.6-35B-A3B | #8 | Accuracy: 96.1 |
| visual-question-answering-vqa-on-realworldqa | Qwen3.6-35B-A3B | #4 | Accuracy: 85.3 |
| visual-question-answering-vqa-on-realworldqa | Qwen3.6-27B | #7 | Accuracy: 84.1 |
| multi-task-language-understanding-on-c-eval | Qwen3.6-27B | #5 | Accuracy: 91.4 |
| multi-task-language-understanding-on-c-eval | Qwen3.6-35B-A3B | #12 | Accuracy: 90.0 |
| multi-task-language-understanding-on-mmlu-redux | Qwen3.6-27B | #5 | Accuracy: 93.5 |
| multi-task-language-understanding-on-mmlu-redux | Qwen3.6-35B-A3B | #7 | Accuracy: 93.3 |
| multimodal-reasoning-on-mathvista | Qwen3.6-27B | #5 | Accuracy: 87.4 |
| multimodal-reasoning-on-mathvista | Qwen3.6-35B-A3B | #7 | Accuracy: 86.4 |
| multimodal-reasoning-on-mmmu | Qwen3.6-27B | #5 | Accuracy: 82.9 |
| multimodal-reasoning-on-mmmu | Qwen3.6-35B-A3B | #7 | Accuracy: 81.7 |
| video-question-answering-on-video-mmmu | Qwen3.6-27B | #5 | Accuracy: 84.4 |
| video-question-answering-on-video-mmmu | Qwen3.6-35B-A3B | #6 | Accuracy: 83.7 |
| vision-language-understanding-on-mmstar | Qwen3.6-27B | #6 | Accuracy: 81.4 |
| vision-language-understanding-on-mmstar | Qwen3.6-35B-A3B | #9 | Accuracy: 80.7 |
| chart-question-answering-on-charxiv-rq | Qwen3.6-27B | #7 | Accuracy: 78.4 |
| chart-question-answering-on-charxiv-rq | Qwen3.6-35B-A3B | #8 | Accuracy: 78.0 |
| optical-character-recognition-ocr-on-ocrbench | Qwen3.6-35B-A3B | #7 | Score: 900 |
| video-question-answering-on-mvbench | Qwen3.6-27B | #8 | Avg.: 75.5 |
| video-question-answering-on-mvbench | Qwen3.6-35B-A3B | #14 | Avg.: 74.6 |
| multi-task-language-understanding-on-mmlu-pro | Qwen3.6-27B | #9 | Accuracy: 86.2 |
| multi-task-language-understanding-on-mmlu-pro | Qwen3.6-35B-A3B | #14 | Accuracy: 85.2 |
| multimodal-reasoning-on-mmmu-pro | Qwen3.6-27B | #9 | Accuracy: 75.8 |
| multimodal-reasoning-on-mmmu-pro | Qwen3.6-35B-A3B | #10 | Accuracy: 75.3 |
| mathematical-reasoning-on-aime-2026 | Qwen3.6-27B | #11 | Accuracy: 94.1 |
| mathematical-reasoning-on-aime-2026 | Qwen3.6-35B-A3B | #15 | Accuracy: 92.7 |
| code-generation-on-livecodebench-v6 | Qwen3.6-27B | #13 | Pass@1: 83.9 |
| code-generation-on-livecodebench-v6 | Qwen3.6-35B-A3B | #18 | Pass@1: 80.4 |
| computer-use-agents-on-androidworld | Qwen3.6-27B | #15 | Success Rate: 70.3 |
| code-generation-on-swe-bench-verified | Qwen3.6-27B (Internal agent scaffold) | #17 | Accuracy: 77.2 |
| code-generation-on-swe-bench-verified | Qwen3.6-35B-A3B (Internal agent scaffold) | #30 | Accuracy: 73.4 |
| code-generation-on-swe-bench-multilingual | Qwen3.6-27B (Internal agent scaffold) | #20 | Accuracy: 71.3 |
| code-generation-on-swe-bench-multilingual | Qwen3.6-35B-A3B (Internal agent scaffold) | #26 | Accuracy: 67.2 |
| multiple-choice-question-answering-mcqa-on-gpqa-diamond | Qwen3.6-27B | #20 | Accuracy: 87.8 |
| multiple-choice-question-answering-mcqa-on-gpqa-diamond | Qwen3.6-35B-A3B | #30 | Accuracy: 86.0 |
| question-answering-on-humanity-s-last-exam-hle | Qwen3.6-27B | #21 | Accuracy: 24.0 |
| question-answering-on-humanity-s-last-exam-hle | Qwen3.6-35B-A3B | #27 | Accuracy: 21.4 |
| code-generation-on-swe-bench-pro | Qwen3.6-27B | #29 | Resolved: 53.5 |
| code-generation-on-swe-bench-pro | Qwen3.6-35B-A3B | #33 | Resolved: 49.5 |