Kuaishou
What benchmarks does Kuaishou mention most?
Here are the 6 benchmarks Kuaishou mentions most in its launch posts.
Source: Labs’ launch posts, read by Benchmark Almanac
SWE-bench Verified LongVideoBench VideoMME v2 (with subtitles) TimeLens-Bench MLVU Video-MME Video-MMMU LiveCodeBench v6 OJBench BFCL V4 tau2-bench VITA-Bench We-Math MathVista DynaMath AIME 2025 AIME 2026 Humanity's Last Exam MMMU MMStar C-Eval IFEval HallusionBench AI2D OCRBench OmniDocBench v1.5 RefCOCO / RefCOCO+ / RefCOCOg PixMo-Count
Source: Labs’ launch posts, system cards and technical reports, read by Benchmark Almanac
Model cards
| Date | Model | Document | Benchmarks |
|---|---|---|---|
| Jul 23, 2026 | | KAT-Coder-V2.5-Dev model card (Hugging Face) | 7 |
| Jul 6, 2026 | | KAT-Coder-V2.5 Technical Report (arXiv 2607.05471) | 6 |
| May 25, 2026 | | Keye-VL-2.0-30B-A3B model card (Hugging Face) | 29 |
| Oct 10, 2025 | | KAT-Dev-72B-Exp model card (Hugging Face) | 1 |
Every Kuaishou model
-
KAT-Coder-V2.5-Dev Jul 2026
-
KAT-Coder-Pro V2.5 Jul 2026
-
Keye-VL-2.0-30B-A3B May 2026
-
KAT-Dev-72B-Exp Oct 2025
-
Keye-VL-1.5-8B Aug 2025
-
Keye-VL-8B-Preview Jun 2025
-
Kling 2.6 Pro
-
Kling v3 Pro