Kimi K3.1 Pro is the flagship model behind Kimi, made by Moonshot AI: a 3-trillion-parameter-class model that reads up to 1 million tokens and is tuned for long coding jobs and knowledge work.
In plain words
It is Kimi K3 with much more post-training, the extra training a model gets after it first learns from data. The Kimi team says it is better than before at coding, at building 3D models and at financial research, and on Terminal-Bench 4.0, a test for coding agents, it scored 65.7%, just under Claude Opus 5.5 at 66.4%. It can also understand images. A Highspeed version runs about two to three times faster. The full weights are due by 4 November 2026.
A quick example
Imagine a developer hands Kimi K3.1 Pro a large code project and asks it to fix a broken feature. The model reads the files, runs the tests in a terminal and keeps trying until they pass, while the developer reviews the final change.
Where you'll meet it
- Kimi.com, the Kimi app, Kimi Work and Kimi Code.
- Apps built on the Kimi API, as kimi-k3.1-pro or kimi-k3.1-pro-highspeed.
- Turning a written description, a sketch or an image into a 3D model.
Related
Sources
- DocsKimi K3.1 Pro Tech Blog: Advancing Open Intelligence, Moonshot AI (Kimi) · read 24 Oct 2026
- DocsModel list, Kimi API Platform · read 24 Oct 2026
- DocsIntroduction, Kimi API Platform · read 24 Oct 2026
Checked by WhatIsAI against the sources above · 24 Oct 2026. A full five-minute explainer with a diagram may follow. Spotted an error? Report it.