Moonshot AI's Kimi Linear 48B A 3B Instruct is a chat model that excels at handling long-context tasks with its hybrid linear attention architecture, achieving significant speedups and reducing memory usage. Notably, it features a refined linear attention mechanism called Kimi Delta Attention (KDA), which optimizes the use of finite-state RNN memory.
Input
Output
Context
1000K
Max Output
-
Parameters
49.1B
Input Modalities
Output Modalities
Estimates based on INT8 quantization. Actual requirements vary by framework and configuration.
Data sourced from official provider APIs and documentation
Last updated: Jun 23, 2026
Automatically route workloads to the right model for every task, every time.