RT Arena.aiBig update: Among open-weight models, Kimi K3 (Max) is #1 in the Agent Arena with +9.75% net-improvement, surpassing GLM-5.2 (Max) at +7.12%, and landed the…
Kimi K3 is now available on @digitalocean 's Serverless Inference! Developers can start building with our most capable model in minutes.DigitalOcean: .@Kimi_Moonshot K3 from Moonshot AI…
Kimi K3 is now available on @nebiustf!Excited to have Nebius as our Day 0 launch partner and bring fast, reliable access to Kimi K3 to more…
Happy to have @FireworksAI_HQ as our day0 launch partner and bring Kimi K3 to more developers.With Fireworks, you can now deploy and fine-tune the 2.8T Kimi…
Excited to have @baseten as our Day 0 launch partner and bring K3 to more users!Baseten's Model APIs provide fast, reliable access to Kimi K3.Baseten: Kimi…
Excited to have @modal as Day 0 launch partner for Kimi K3!They trained a custom DFlash speculator for K3's architecture, delivering faster inference with no quality…
We've open-sourced MoonEP, our high-performance communication library for distributed MoE workloads. Built to make expert-parallel communication more efficient at scale, MoonEP helps reduce communication overhead in…
We've open-sourced AgentENV in collaboration with kvcache-ai.AgentENV is a distributed system for running agent environments at scale. Its components power agentic RL training for Kimi K3,…
We've open-sourced FlashKDA, our high-performance CUTLASS-based implementation of Kimi Delta Attention kernels.It delivers 1.72×–2.22× prefill speedup over the flash-linear-attention baseline on H20, and works as a…
RT Artificial AnalysisKimi K3 is second only to Fable 5 on AA-Briefcase, our agentic knowledge work benchmark, but costs more than Opus 4.8 to run while…