Coding-agent users get a capable open model at under $1 per million input tokens, with self-hosting as an option.
Kenning's take
What it does
Kimi K2.6 is an open-weight model from Moonshot AI aimed at long coding runs, front-end design from prompts, and agents that keep working in the background. Its Agent Swarm scales to 300 sub-agents across 4,000 steps, up from 100 in K2.5. It is available on Kimi.com, the Kimi app, Kimi Code and the Kimi API, and as weights on Hugging Face.
- Long-horizon coding in Rust, Go and Python
- Front-end generation from prompts
- Agent Swarm up to 300 sub-agents
- Thinking and non-thinking modes with image and video input
- 256K token context
Who it suits. Developers and teams who run coding agents or long background tasks.
Key facts
- Category
- AI models
- Part of
- Kimi
- Made by
- Moonshot AI
- Based in
- China
- Released
- 20 April 2026
- How to get it
- on Kimi.com, the Kimi app, Kimi Code and the Kimi API (kimi-k2.6), plus open weights on Hugging Face under a Modified MIT licence
- Works on
- Web, iOS, Android, API, self-hosted
- Free plan
- Weights free to download; chat access on Kimi.com
- API price
- $0.95 per million input tokens ($0.16 cached), $4 per million output
Reviews and write-ups
FAQ
When was Kimi K2.6 released?
Moonshot announced it on 20 April 2026, according to press reports.
How much does the Kimi K2.6 API cost?
$0.95 per million input tokens, $0.16 for cached input, and $4 per million output tokens.
Is Kimi K2.6 open source?
The weights are on Hugging Face under a Modified MIT licence.
Facts from the maker's own pages and the sources listed. Checked 2026-10-11.
Similar tools

Kimi K2
Moonshot AI's first trillion-parameter open-weight model, built for coding and tool use

Kimi K2 Thinking
Reasoning version of Kimi K2 that thinks step by step between hundreds of tool calls

Kimi K2.5
Moonshot AI's open multimodal model that added image and video input and agent swarms to Kimi

Kimi K2.7 Code
Moonshot AI's open coding model, built on K2.6 and using about 30% fewer thinking tokens