> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.mixlayer.com/kimi-k2-7-code/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.mixlayer.com/_mcp/server.

# Kimi K2.7 Code

> Guidance for coding and agentic workloads with Kimi K2.7 Code

Kimi K2.7 Code is a coding-specialized model from Moonshot AI. It is designed for long-horizon software engineering, tool use, and agentic repository work.

## Model details

* **Identifier:** `moonshotai/kimi-k2.7-code`
* **Context window:** 256K tokens
* **Capabilities:** Text · Reasoning · Tools

## Best for

* Exploring and modifying large repositories
* Multi-file implementation and refactoring
* Long-running coding agents
* Debugging, test repair, and terminal-based workflows

## Platform defaults

When sampling parameters are omitted, Mixlayer applies:

```json
{
  "temperature": 1.0,
  "top_p": 0.95
}
```

## Reasoning behavior

Kimi K2.7 Code uses always-on thinking. Budget enough output tokens for both reasoning and the final answer, especially for long agentic tasks.

Use clear objectives, provide relevant repository context, and expose focused tools. For long tasks, let the model inspect the codebase before asking it to implement changes.

> **Info**
>
> Call `GET /v1/models` to confirm current availability and limits for your API
> key.

See the [official Kimi K2.7 Code model card](https://huggingface.co/moonshotai/Kimi-K2.7-Code) for upstream architecture and behavior.

> Guidance for coding and agentic workloads with Kimi K2.7 Code