Quick Start

Kimi Code CLI is an AI agent that runs in the terminal, helping you complete software development tasks and terminal operations. It can read and edit code, execute shell commands, search and fetch web pages, and autonomously plan and adjust actions during execution.

Kimi Code CLI is suited for:

• Writing and modifying code: Implementing new features, fixing bugs, refactoring code
• Understanding projects: Exploring unfamiliar codebases, answering architecture and implementation questions
• Automating tasks: Batch processing files, running builds and tests, executing scripts

Kimi Code CLI supports the following usage modes:

• Interactive CLI (kimi): Chat with AI in the terminal using natural language to describe tasks or execute shell commands directly
• Browser UI (kimi web): Open a graphical interface in your local browser, supporting session management, file references, code highlighting, and more
• Agent integration (kimi acp): Run as a service and integrate into IDEs and other local agent clients via the Agent Client Protocol

If you encounter issues or have suggestions, feel free to provide feedback via GitHub Issues.

Before You Start

• Operating system: macOS, Linux, or Windows (via PowerShell)
• Kimi account: Requires a Kimi membership subscription, or a callable API key

Installation

Run the installation script to complete the installation. The script will first install uv (a Python package manager), then install Kimi Code CLI via uv:

# Linux / macOS
curl -LsSf https://code.kimi.com/install.sh | bash

# Windows (PowerShell)
Invoke-RestMethod https://code.kimi.com/install.ps1 | Invoke-Expression

Verify the installation:

kimi --version

Due to macOS security checks (Gatekeeper), the first run of the kimi command may take longer. You can add your terminal application in "System Settings → Privacy & Security → Developer Tools" to speed up subsequent launches.

If you already have uv installed, you can also run:

uv tool install --python 3.13 kimi-cli

Kimi Code CLI supports Python 3.12–3.14, but Python 3.13 is recommended for best compatibility.

First Run

Start and Log In

Start Kimi Code CLI in the project directory where you want to work:

cd your-project
kimi

On first launch, enter /login to configure the API source:

/login

We recommend selecting the Kimi Code platform, which automatically opens the browser for OAuth authorization; selecting other platforms requires entering an API key. After configuration, it automatically saves and reloads. See the Providers and Models configuration documentation for details.

Step 1: Ask a Question

Ask a question in natural language to quickly understand the project:

What is the overall architecture of this project? Where is the entry file?

Kimi Code CLI will automatically search and read relevant files, then provide an answer.

Step 2: Make a Code Change

Try letting Kimi Code CLI modify code:

Add a "Quick Start" section to the README, including installation and running steps

Kimi Code CLI will show a diff and ask for confirmation before modifying files—you can approve, reject, or enter feedback to adjust direction. It won't change your code without permission.

Step 3: Execute a Command

Kimi Code CLI can also run shell commands and analyze the results:

Run the tests and fix any failing cases

At this point, you've experienced three core capabilities: asking questions, modifying code, and executing commands.

If the project doesn't have an AGENTS.md file, you can run the /init command to have Kimi Code CLI analyze the project and generate this file, helping the AI better understand the project structure and conventions.

Common Commands Cheat Sheet

Command	Description
kimi	Start an interactive conversation
kimi web	Open the browser graphical interface
/login	Configure or switch API source
/usage	View remaining quota and limits
/help	View all commands and shortcuts
Ctrl-J	Newline (without submitting)
Ctrl-C / Ctrl-D	Interrupt current operation / Exit

For the full command list, please refer to Slash Commands and Keyboard Shortcuts.

FAQ

Why do I get an authentication failure after entering my API Key?

First, confirm that your Key and Base URL belong to the same platform. api.kimi.com and api.moonshot.cn are two completely independent account systems; API Keys are not interchangeable:

Platform	Base URL	Billing	Key Creation Portal
Kimi Code	OpenAI Compatible: https://api.kimi.com/coding/v1 Anthropic Compatible: https://api.kimi.com/coding/	Kimi membership subscription (with quota)	www.kimi.com/code/console
Kimi Platform	https://api.moonshot.cn/v1	Pay-as-you-go	platform.kimi.ai

What to do: Go back to the corresponding platform's console to recreate a Key, and ensure the base_url (also called Entrypoint or Endpoint) configured in your tool matches the Key's platform.

kimi command not found after installation

The installation script adds kimi to PATH, but you need to restart the terminal or run source ~/.bashrc (or source ~/.zshrc) for it to take effect. If it's still not found, check whether ~/.local/bin is in your PATH.

Browser doesn't pop up after /login

If you're on a remote server or in a headless environment, /login will display a URL; manually copy it to your browser to complete authorization.

For more issues, please refer to FAQ.

Upgrade and Uninstall

Upgrade to the latest version:

uv tool upgrade kimi-cli --no-cache

Uninstall Kimi Code CLI:

uv tool uninstall kimi-cli

Next Steps

• Core Operations – Learn interaction input, context management, working modes, and other interaction tips
• Configuration Files — Customize models, behavior, and tool permissions
• MCP Integration — Connect external tools and data sources to extend capabilities