Moonshot AI released Kimi K3, an open frontier model that swept leaderboards, featuring a hybrid attention mechanism with Kimi Delta Attention (KDA), a linear attention layer derived from DeltaNet and…
Moonshot AI open-sourced MoonEP, an Expert Parallelism communication library for distributed Mixture-of-Experts workloads, under an MIT license. The library guarantees every rank receives exactly S × …
Moonshot AI released Kimi K3, a 2.8T-parameter open-weights MoE model with 104B active parameters, 896 experts, 1M-token context, and native visual understanding, which has been independently validate…
Moonshot AI released the model weights and technical report for Kimi K3, a 2.8-trillion-parameter Mixture-of-Experts model with native visual understanding and a 1-million-token context window, along …