LLxprt Code

   

LLxprt Code is a powerful fork of Google's Gemini CLI, enhanced with multi-provider support and improved theming. We thank Google for their excellent foundation and will continue to track and merge upstream changes as long as practical.

Key Features

• Multi-Provider Support: Direct access to OpenAI (o3), Anthropic (Claude), Google Gemini, plus OpenRouter, Fireworks, and local models
• Enhanced Theme Support: Beautiful themes applied consistently across the entire tool
• Full Gemini CLI Compatibility: All original features work seamlessly, including Google authentication via /auth
• Local Model Support: Run models locally with LM Studio, llama.cpp, or any OpenAI-compatible server
• Flexible Configuration: Switch providers, models, and API keys on the fly
• Advanced Settings & Profiles: Fine-tune model parameters, manage ephemeral settings, and save configurations for reuse. Learn more →

With LLxprt Code you can:

• Query and edit large codebases with any LLM provider
• Generate new apps from PDFs or sketches, using multimodal capabilities
• Use local models for privacy-sensitive work
• Switch between providers seamlessly within a session
• Leverage all the powerful tools and MCP servers from Gemini CLI
• Use tools and MCP servers to connect new capabilities, including media generation with Imagen, Veo or Lyria
• Ground your queries with the Google Search tool when using Gemini
• Enjoy a beautifully themed interface across all commands

Quickstart

You have two options to install LLxprt Code.

With Node

1. Prerequisites: Ensure you have Node.js version 20 or higher installed.

2. Install LLxprt Code:

   npm install -g @vybestack/llxprt-code

   Or run directly with npx:

   npx https://github.com/acoliver/llxprt-code

Common Configuration Steps

3. Run and configure:

   llxprt

   • Pick a beautiful theme
   • Choose your provider with /provider (defaults to Gemini)
   • Set up authentication as needed

Provider Configuration

Using OpenAI

Direct access to o3, o1, GPT-4.1, and other OpenAI models:

1. Get your API key from OpenAI
2. Configure LLxprt Code:

   /provider openai
   /key sk-your-openai-key-here
   /model o3-mini
   

Using Anthropic

Access Claude Sonnet 4, Claude Opus 4.1, and other Anthropic models:

Option 1: Log in with Anthropic to use your Claude Pro or Max account

Use OAuth authentication to access Claude with your existing Claude Pro or Max subscription:

1. Select the Anthropic provider:

   /provider anthropic
   

2. Authenticate with your Claude account:

   /auth
   

3. Your browser will open to the Claude authentication page
4. Log in and authorize LLxprt Code
5. Copy the authorization code shown and paste it back in the terminal
6. You're now using your Claude Pro/Max account!

Option 2: Use an API Key

1. Get your API key from Anthropic
2. Configure:

   /provider anthropic
   /key sk-ant-your-key-here
   /model claude-sonnet-4-20250115
   

Using Qwen

Access Qwen3-Coder-Pro and other Qwen models for free:

Option 1: Log in with Qwen (FREE)

Use OAuth authentication to access Qwen with your free account:

/auth qwen enable

This enables OAuth for Qwen. When you send your first message, your browser will automatically open to the Qwen authentication page. Log in and authorize LLxprt Code - the authentication will complete automatically and your request will be processed. You're now using Qwen3-Coder-Pro for free!

Option 2: Use an API Key

For advanced users who need API access:

1. Get your API key from Qwen
2. Configure:

   /provider qwen
   /key your-qwen-api-key
   /model qwen3-coder-pro
   

Using Cerebras Code Max/Pro

Access Cerebras Code Max/Pro plan with the powerful qwen-3-coder-480b model:

1. Get your API key from Cerebras
2. Configure LLxprt Code:

   /provider openai
   /baseurl https://api.cerebras.ai/v1
   /key your-cerebras-api-key
   /model qwen-3-coder-480b
   

For optimal performance with this model, consider setting a high context limit:

/set context-limit 100000

Using Local Models

Run models locally for complete privacy and control. LLxprt Code works with any OpenAI-compatible server.

Example with LM Studio:

1. Start LM Studio and load a model (e.g., Gemma 3B)
2. In LLxprt Code:

   /provider openai
   /baseurl http://127.0.0.1:1234/v1/
   /model gemma-3b-it
   

Example with llama.cpp:

1. Start llama.cpp server: ./server -m model.gguf -c 2048
2. In LLxprt Code:

   /provider openai
   /baseurl http://localhost:8080/v1/
   /model local-model
   

List available models:

/model

This shows all models available from your current provider.

Using OpenRouter

Access 100+ models through OpenRouter:

1. Get your API key from OpenRouter
2. Configure LLxprt Code:

   /provider openai
   /baseurl https://openrouter.ai/api/v1/
   /keyfile ~/.openrouter_key
   /model qwen/qwen3-coder
   /profile save qwen3-coder
   

Using Fireworks

For fast inference with popular open models:

1. Get your API key from Fireworks
2. Configure:

   /provider openai
   /baseurl https://api.fireworks.ai/inference/v1/
   /key fw_your-key-here
   /model accounts/fireworks/models/llama-v3p3-70b-instruct
   

Using xAI (Grok)

Access Grok models through xAI's API:

1. Get your API key from xAI

2. Configure using command line:

   llxprt --provider openai --baseurl https://api.x.ai/v1/ --model grok-3 --keyfile ~/.mh_key

   Or configure interactively:

   /provider openai
   /baseurl https://api.x.ai/v1/
   /model grok-3
   /keyfile ~/.mh_key
   

3. List available Grok models:

   /model
   

Using Google Gemini

You can still use Google's services:

1. With Google Account: Use /auth to sign in
2. With API Key:

   export GEMINI_API_KEY="YOUR_API_KEY"

   Or use /key YOUR_API_KEY after selecting the gemini provider

Managing API Keys

• Set key for current session: /key your-api-key
• Load key from file: /keyfile ~/.keys/openai.txt
• Environment variables: Still supported for all providers

Examples

Start a new project:

cd new-project/
llxprt
> Create a Discord bot that answers questions using a FAQ.md file I will provide

Work with existing code:

git clone https://github.com/acoliver/llxprt-code
cd llxprt-code
llxprt
> Give me a summary of all the changes that went in yesterday

Use a local model for sensitive code:

llxprt
/provider openai
/baseurl http://localhost:1234/v1/
/model codellama-7b
> Review this code for security vulnerabilities

Control and Shortcuts: Settings and Profiles

LLxprt Code provides powerful configuration options through model parameters and profiles:

# Fine-tune model behavior
/set modelparam temperature 0.8
/set modelparam max_tokens 4096

# Configure context handling
/set context-limit 100000
/set compression-threshold 0.7

# Save your configuration
/profile save my-assistant

# Load it later
llxprt --profile-load my-assistant

See the complete settings documentation for all configuration options.

Key Features

Code Understanding & Generation

• Query and edit large codebases
• Generate new apps from PDFs, images, or sketches using multimodal capabilities
• Debug issues and troubleshoot with natural language

Automation & Integration

• Automate operational tasks like querying pull requests or handling complex rebases
• Use MCP servers to connect new capabilities
• Run non-interactively in scripts for workflow automation

Advanced Capabilities

• Ground your queries with built-in search capabilities for real-time information
• Conversation checkpointing to save and resume complex sessions
• Custom context files (LLXPRT.md) to tailor behavior for your projects

GitHub Integration (Experimental)

Note: GitHub Actions integration is currently experimental and disabled by default in LLxprt Code. This feature is under development and may not be fully functional.

For potential future GitHub workflow integration:

• Pull Request Reviews: Automated code review with contextual feedback and suggestions
• Issue Triage: Automated labeling and prioritization of GitHub issues based on content analysis
• On-demand Assistance: Mention capabilities in issues and pull requests for help with debugging, explanations, or task delegation
• Custom Workflows: Build automated, scheduled and on-demand workflows tailored to your team's needs

Customizing AI Behavior with Prompts

LLxprt Code features a sophisticated prompt configuration system that allows you to customize the AI's behavior for different providers, models, and use cases. You can:

• Create custom system prompts for specific tasks
• Override provider-specific behaviors
• Add environment-aware instructions
• Customize tool usage guidelines

Learn more in the Prompt Configuration Guide.

Next steps

• Learn how to contribute to or build from the source.
• Explore the available CLI Commands.
• If you encounter any issues, review the troubleshooting guide.
• For more comprehensive documentation, see the full documentation.
• Migrating from Gemini CLI? Check out our tips for Gemini CLI users.
• Take a look at some popular tasks for more inspiration.
• Check out our Official Roadmap

Provider Commands Reference

• /provider - List available providers or switch provider
• /model - List available models or switch model
• /baseurl - Set custom API endpoint
• /key - Set API key for current session
• /keyfile - Load API key from file
• /auth - Authenticate with Google (for Gemini), Anthropic (for Claude), or Qwen

Documentation

• Architecture Overview - How llxprt works
• IDE Integration - VS Code companion
• Sandboxing & Security - Safe execution environments
• Enterprise Deployment - Docker, system-wide config
• Telemetry & Monitoring - Usage tracking
• Tools API Development - Create custom tools

Troubleshooting

See the troubleshooting guide if you encounter issues.

Popular tasks

Explore a new codebase

Start by cding into an existing or newly-cloned repository and running llxprt.

> Describe the main pieces of this system's architecture.

> What security mechanisms are in place?

> Provide a step-by-step dev onboarding doc for developers new to the codebase.

> Summarize this codebase and highlight the most interesting patterns or techniques I could learn from.

> Identify potential areas for improvement or refactoring in this codebase, highlighting parts that appear fragile, complex, or hard to maintain.

> Which parts of this codebase might be challenging to scale or debug?

> Generate a README section for the [module name] module explaining what it does and how to use it.

> What kind of error handling and logging strategies does the project use?

> Which tools, libraries, and dependencies are used in this project?

Work with your existing code

> Implement a first draft for GitHub issue #123.

> Help me migrate this codebase to the latest version of Java. Start with a plan.

Automate your workflows

Use MCP servers to integrate your local system tools with your enterprise collaboration suite.

> Make me a slide deck showing the git history from the last 7 days, grouped by feature and team member.

> Make a full-screen web app for a wall display to show our most interacted-with GitHub issues.

Interact with your system

> Convert all the images in this directory to png, and rename them to use dates from the exif data.

> Organize my PDF invoices by month of expenditure.

Uninstall

Head over to the Uninstall guide for uninstallation instructions.

Privacy and Terms

LLxprt Code does not collect telemetry by default. Your privacy is important to us.

When using Google's services through LLxprt Code, you are bound by Google's Terms of Service and Privacy Notice. Other providers have their own terms that apply when using their services.