Looking forward to running longer tasks and expanding LLM usage for product flows more effectively, rather than just for development.
Also future is self-hosted ๐
Introducing Kimi K3: Open Frontier Intelligence
๐น 2.8 Trillion Parameters, 1 Million Context, Native Multimodal
๐น Kimi Delta Attention enables up to 6.3x faster decoding in million-token contexts
๐น Attention Residuals deliver ~25% higher training efficiency at <2% additional cost
๐น Built for long-horizon agentic coding and self-evolving workflows
Kimi K3 is now live on on Kimi Work, Kimi Code, and the Kimi API.
Open Weights by July 27, 2026.
๐ API:
๐ Tech blog: