Moonshot releases Kimi K3, a 2.8T open-weight multimodal model with a 1M-token context

cyb3rops · x · 2026-07-27

Moonshot AI has released Kimi K3 on Hugging Face as an open-weight, native multimodal agentic model with 2.8T parameters and a 1M-token context window.

The model card says Kimi K3 is built on Kimi Delta Attention and Attention Residuals, activates 16 of 896 experts under a Stable LatentMoE design, and claims roughly 2.5× better scaling efficiency than Kimi K2. Moonshot positions it for long-horizon coding, knowledge work, reasoning, and multimodal tasks such as vision-in-the-loop workflows, dashboards, motion design, and video editing.

Related event: Moonshot Open-Sources Kimi K3: 2.8T Params and $20M Commercial Threshold(148 posts)→

Original post →

More from Models

Models channel →