Kimi K3 Tops Hugging Face Trending With 2.8T Parameters

Moonshot AI’s Kimi K3 is the most-trended model on Hugging Face this week: 2.8 trillion total parameters with 104 billion active, routing 16 of 896 experts per token across 93 layers of Kimi Delta Attention and gated MLA. Context is just over one million tokens, and the release handles text, image and video through a 401M-parameter MoonViT-V2 encoder. Vendor benchmarks include 93.5 on GPQA Diamond, 88.3 on Terminal-Bench 2.1 and 84.8 on OSWorld-Verified. Weights ship quantisation-aware in MXFP4 under Moonshot’s own licence — open weights, but not an OSI-approved licence.