Moonshot AI Kimi K3

The Kimi K3 family from Moonshot AI introduces next-generation multimodal Mixture-of-Experts model built for frontier-scale reasoning, coding, and agentic workflows. As the world's first open 3T-class model family, it combines native multimodal understanding, million-token context, and advanced attention architectures for efficient long-horizon performance.

1 variant

Kimi K3

Moonshot AI

Params: 2.8T

Kimi K3 is a frontier-scale native multimodal Mixture-of-Experts model designed for long-horizon reasoning, agentic coding, and large-scale knowledge workflows. It is the world's first open 3T-class model, featuring 2.8T total parameters with 104B activated, built on a 93-layer architecture with Kimi Delta Attention (KDA), Gated MLA, and Stable LatentMoE. The model uses a 7,168 hidden size, 96 attention heads, and 896 routed experts, activating 16 experts per token alongside 2 shared experts. Powered by MoonViT-V2 for native vision understanding and supporting a 1M-token context window, Kimi K3 is optimized for multimodal reasoning, long-context coding, advanced agentic workflows, and frontier-scale research applications

Text GenerationInstruction Following+8

Runs onMI355X+

Runs on

Moonshot AI Kimi K3 Models | Vultr Docs