AI
Moonshot publishes Kimi K3.1 Pro weights at 2.8 trillion parameters
Moonshot AI has published the full weights of Kimi K3.1 Pro on Hugging Face, a 2.8-trillion-parameter mixture-of-experts model that activates about 104 billion parameters per token with a context window of roughly one million tokens. The checkpoint ships in MXFP4 across 96 shards totalling about 1.56 terabytes under a custom licence, and the model card recommends vLLM, SGLang or TokenSpeed to serve it. It handles text and image input.