| visual-question-answering-vqa-on-blink | Kimi K2.5 | #1 | Accuracy: 78.9 |
| mathematical-reasoning-on-hmmt-feb-2025 | Kimi K2.5 | #2 | Accuracy: 95.4 |
| multimodal-reasoning-on-mathvista | Kimi K2.5 | #2 | Accuracy: 90.1 |
| multimodal-reasoning-on-mmmu | Kimi K2.5 | #2 | Accuracy: 84.3 |
| optical-character-recognition-ocr-on-ocrbench | Kimi K2.5 | #2 | Score: 923 |
| multimodal-reasoning-on-zerobench | Kimi K2.5 | #3 | Accuracy: 9 |
| video-question-answering-on-video-mmmu | Kimi K2.5 | #3 | Accuracy: 86.6 |
| visual-question-answering-vqa-on-simplevqa | Kimi K2.5 | #3 | Accuracy: 71.2 |
| long-context-understanding-on-longbench-v2 | Kimi K2.5 | #4 | Accuracy: 61.0 |
| video-question-answering-on-mmvu | Kimi K2.5 | #4 | Accuracy: 80.4 |
| code-generation-on-scicode | Kimi K2.5 | #5 | Accuracy: 48.7 |
| multi-task-language-understanding-on-mmlu-pro | Kimi K2.5 | #5 | Accuracy: 87.1 |
| multimodal-reasoning-on-mmmu-pro | Kimi K2.5 | #5 | Accuracy: 78.5 |
| video-question-answering-on-lvbench | Kimi K2.5 | #5 | Accuracy: 75.9 |
| video-question-answering-on-video-mme-with-subtitles | Kimi K2.5 | #5 | Accuracy: 87.4 |
| code-generation-on-terminal-bench-2-0-terminus-2 | Kimi K2.5 (Terminus-2) | #6 | Accuracy: 50.8 |
| mathematical-reasoning-on-aime-2025 | Kimi K2.5 | #6 | Accuracy: 96.1 |
| multimodal-reasoning-on-mathvision | Kimi K2.5 | #7 | Accuracy: 84.2 |
| multimodal-reasoning-on-babyvision | Kimi K2.5 | #8 | Accuracy: 36.5 |
| chart-question-answering-on-charxiv-rq | Kimi K2.5 | #9 | Accuracy: 77.5 |
| code-generation-on-livecodebench-v6 | Kimi K2.5 | #11 | Pass@1: 85.0 |
| question-answering-on-humanity-s-last-exam-hle | Kimi K2.5 | #12 | Accuracy: 30.1 |
| code-generation-on-swe-bench-multilingual | Kimi K2.5 (Internal agent scaffold) | #18 | Accuracy: 73.0 |
| code-generation-on-swe-bench-verified | Kimi K2.5 (Internal agent scaffold) | #18 | Accuracy: 76.8 |
| multiple-choice-question-answering-mcqa-on-gpqa-diamond | Kimi K2.5 | #21 | Accuracy: 87.6 |
| deep-research-agents-on-browsecomp | Kimi K2.5 (w/ ctx manage) | #23 | Accuracy: 74.9 |
| deep-research-agents-on-browsecomp | Kimi K2.5 | #38 | Accuracy: 60.6 |
| code-generation-on-swe-bench-pro | Kimi K2.5 | #32 | Resolved: 50.7 |