Repolode
MoonshotAI

MoonshotAI/kimi-code

An AI coding agent for the terminal that reads and edits code, runs shell commands, and works with Kimi and compatible models.

Open-source alternative to:Claude Code

Stars
8,154
Forks
1,355
Open issues
1,782
Last commit
23 hours ago

What is it?

Kimi Code CLI is an AI coding agent that runs in the terminal. It can read and edit code, run shell commands, search files, fetch web pages, and pick its next step based on the feedback it gets. It works out of the box with Moonshot AI's Kimi models and can be configured to use other compatible providers.

The tool is distributed as a single binary installed with one shell command, so Node.js does not need to be set up beforehand. Sessions run in an interactive terminal UI, with approvals for actions, session history, and built-in subagents (coder, explore, plan) that work in isolated contexts. Users can also add MCP servers, skills and data sources from a marketplace or GitHub repos, and attach lifecycle hooks that run local commands at key points.

The CLI can also serve as a backend for editors and IDEs that support the Agent Client Protocol, such as Zed and JetBrains, via the `kimi acp` subcommand. On Windows it requires Git for Windows because it uses Git Bash as its shell environment.

Key features

  • 01Single-command install with no Node.js setup required.
  • 02Reads, edits, and searches code and runs shell commands from the terminal.
  • 03Accepts video input such as screen recordings as part of a chat.
  • 04Configures MCP servers through conversation with /mcp-config.
  • 05Runs built-in coder, explore, and plan subagents in isolated contexts.
  • 06Supports lifecycle hooks and ACP integration with Zed and JetBrains.

Use cases

  • Exploring an unfamiliar codebase and asking how its main directories fit together.
  • Making code changes and running tests or commands from the terminal.
  • Driving an agent session from inside Zed or a JetBrains IDE.
  • Extending the agent with MCP servers and skills for custom data sources.

Quick start

  1. 1

    Run the official install script from the README (macOS, Linux, or Windows PowerShell)

  2. 2

    Verify the install in a new shell session

    kimi --version
  3. 3

    Open your project folder and start the interactive UI

    kimi
  4. 4

    Run /login inside Kimi Code CLI with Kimi Code OAuth or a Moonshot AI API key

Things to know

  • On Windows it requires Git for Windows to be installed first.
  • Login needs either a Kimi Code OAuth account or a Moonshot AI Open Platform API key.

In collections

Topics

FAQ

What is kimi-code?
An AI coding agent for the terminal that reads and edits code, runs shell commands, and works with Kimi and compatible models. Kimi Code CLI is an AI coding agent that runs in the terminal. It can read and edit code, run shell commands, search files, fetch web pages, and pick its next step based on the feedback it gets. It works out of the box with Moonshot AI's Kimi models and can be configured to use other compatible providers.
Is kimi-code free to use? What license does it have?
kimi-code is open source under the MIT license. Check the license terms before using it in commercial products.
Is kimi-code still maintained?
Actively maintained. The last commit to kimi-code was on Nov 6, 2026.
What can I use kimi-code for?
Exploring an unfamiliar codebase and asking how its main directories fit together. Making code changes and running tests or commands from the terminal. Driving an agent session from inside Zed or a JetBrains IDE. Extending the agent with MCP servers and skills for custom data sources.
How do I install kimi-code?
Run the official install script from the README (macOS, Linux, or Windows PowerShell) Verify the install in a new shell session: kimi --version Open your project folder and start the interactive UI: kimi Run /login inside Kimi Code CLI with Kimi Code OAuth or a Moonshot AI API key
Where can I get kimi-code?
The source code and installation instructions are on GitHub: https://github.com/MoonshotAI/kimi-code

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects