Overview
Kimi Code’s K2.8 preview model, with coding and agent capability close to K3, more efficient thinking than K2.7 Code, image/video input, and up to 1M context.
Kimi Code’s K2.8 preview model, with coding and agent capability close to K3, more efficient thinking than K2.7 Code, image/video input, and up to 1M context.
Kimi Code fully rolled out K2.8 Preview on 2026-09-11; the Model ID remains `kimi-for-coding`, it is available to all members, supports low, high, and max thinking (default max), and routes no-thinking requests to the K2.8 non-thinking version.
Only published specifications are shown.
Kimi Code’s K2.8 preview model, with coding and agent capability close to K3, more efficient thinking than K2.7 Code, image/video input, and up to 1M context.
K2.8 Preview upgrades Kimi Code’s default coding route with more efficient reasoning while preserving the existing Model ID, allowing the preview to replace kimi-for-coding without client changes.
Kimi K2.8 Preview fully rolled out in Kimi Code; the Model ID `kimi-for-coding` is unchanged, with low, high, and max thinking and up to 1M context.
View sourceSelect a card to view the related model.
Copy this template, add the field, proposed value, and primary source, and share it with the archive maintainer.