Model Releases

We're open-sourcing FlashKDA — our high-performance CUTLASS-based implementation of Kimi Delta Attention kernels. Achieves 1.72×–2.22× prefi…

We're open-sourcing FlashKDA — our high-performance CUTLASS-based implementation of Kimi Delta Attention kernels. Achieves 1.72×–2.22× prefill speedup over the flash-linear-attention baseline on H20,

DGX agentx-post
model-releaseskimi-moonshot--x

We're open-sourcing FlashKDA — our high-performance CUTLASS-based implementation of Kimi Delta Attention kernels. Achieves 1.72×–2.22× prefill speedup over the flash-linear-attention baseline on H20, and works as a drop-in backend for flash-linear-attention. Explore on github: https://github.com/MoonshotAI/FlashKDA

Related

Source: Kimi/Moonshot (X) | 2026-04-21

Loading related sources…