Package Exports

cli-llm-switcher/package.json

This package does not declare an exports field, so the exports above have been automatically detected and optimized by JSPM instead. If any package subpath is missing, it is recommended to post an issue to the original package (cli-llm-switcher) to support the "exports" field. If that is not possible, create a JSPM override to customize the exports field for this package.

Readme

cli-llm-switcher

A command-line tool for seamlessly switching between multiple LLM providers

Seamlessly switch between LLM providers without affecting your native Claude Code setup. Isolated configuration with official best practices from model providers.

💡 Why LLM Switcher?

🔒 Isolated Configuration Environment

Zero impact on native Claude Code: Your original Claude setup remains untouched
Per-session provider switching: Each terminal session can use different providers

🎯 Official Best Practices

Provider-recommended configurations: Following official integration guidelines from each provider
Claude Code dual-model configuration: Main model for conversation/planning/code generation/complex reasoning, fast model (Claude uses Haiku e.g. 3.5 Haiku) for file search/syntax checking and auxiliary tasks - intelligently optimizing performance and cost

📋 Prerequisites

Before installing, ensure you have Node.js (v16 or later) installed:

Install Node.js

Option 1 (Recommended): Use nvm for easy Node.js management

# Install nvm: https://github.com/nvm-sh/nvm#install--update-script
curl -o- https://raw.githubusercontent.com/nvm-sh/nvm/v0.39.0/install.sh | bash
# Restart terminal, then install latest Node.js
nvm install node
nvm use node

Option 2: Download from nodejs.org (choose LTS version)

Verify installation:

node --version  # Should show v16.0.0 or higher
npm --version   # Should show npm version

🚀 Installation

# Install globally via npm
npm install -g cli-llm-switcher

# Verify installation
lms --version

# Configure API keys
lms config
# Follow prompts to enter your API key

# Start using with Claude Code or other tools
lms run claude

🤖 Provider Integration

DeepSeek

Model Configuration

Main Model: deepseek-chat
Fast Model: deepseek-chat

Get Your API Key

Platform: https://platform.deepseek.com/

AlibabaCloud (International)

Model Configuration

Main Model: qwen3-coder-plus
Fast Model: qwen3-coder-flash

Get Your API Key

🌍 International: https://modelstudio.console.alibabacloud.com/

AlibabaCloud (China)

Model Configuration

Main Model: qwen3-coder-plus
Fast Model: qwen3-coder-flash

Get Your API Key

🇨🇳 China: https://bailian.console.aliyun.com/

Moonshot AI

Model Configuration

Main Model: kimi-k2-0905-preview
Fast Model: kimi-k2-0905-preview

Get Your API Key

Platform: https://platform.moonshot.ai/

Zhipu GLM

Model Configuration

Main Model: glm-4.5
Fast Model: glm-4.5-air

Get Your API Key

🇨🇳 China: https://bigmodel.cn/
🌍 International: https://z.ai/model-api

Grok Code Fast 1

Model Configuration

Main Model: grok-code-fast-1
Fast Model: grok-code-fast-1

Get Your API Key

Platform: https://console.x.ai

LongCat

Model Configuration

Main Model: LongCat-Flash-Chat
Fast Model: LongCat-Flash-Chat

Get Your API Key

Platform: https://longcat.chat/platform/api_keys

Uninstallation

Basic Uninstall (keeps configuration)

npm uninstall -g cli-llm-switcher

Complete Uninstall (removes everything)

Note: Run lms status to see the configuration directory path before uninstalling.

macOS/Linux:

npm uninstall -g cli-llm-switcher
rm -rf ~/.llm-switch

Windows (PowerShell):

npm uninstall -g cli-llm-switcher
Remove-Item -Recurse -Force "$env:USERPROFILE\.llm-switch"

Windows (Command Prompt):

npm uninstall -g cli-llm-switcher
rmdir /s /q "%USERPROFILE%\.llm-switch"

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

This project is licensed under the MIT License - see the LICENSE file for details.

References

Official provider configuration guides for Claude Code integration:

Need help? Check the complete documentation for detailed guides and troubleshooting.