One workflow, provider options
Configure supported providers and model choices for your setup. Availability and capabilities depend on the provider and selected model.
Independent open-source project
Your coding workflow.
Your model choices.
Free Claude Code helps route coding-agent requests through supported hosted or local model providers, with configuration options for different workflows.
Provider availability, pricing, limits, and terms are controlled by each provider and can change.
terminal — quick start
$ uv tool install --force git+https://github.com/Alishahryar1/free-claude-code.git
# Install the package
$ fcc-server
✓ Start the local server
Configure your chosen provider before use.
›
Open source
Provider flexibility
Hosted or local models
Community contributions
01 / Overview
A flexible layer for your
coding-agent setup.
Free Claude Code is a community project that provides an Anthropic-compatible proxy and integrations for supported model providers. It is designed to let you configure model routing without changing the coding-agent workflow more than necessary.
Hosted or local
Use supported hosted services or compatible local runtimes such as LM Studio or llama.cpp, where configured.
Built in the open
Review the source, report issues, and contribute improvements through the public GitHub repository.
02 / Features
Tools for a more configurable workflow.
Feature descriptions below are based on the public repository README. Check the repository for the current implementation and requirements.
Provider integrations
Route requests through supported providers, including NVIDIA NIM, OpenRouter, DeepSeek, LM Studio, and llama.cpp. Check the README for the current list.
Per-model routing
Configure provider/model choices for model names such as Opus, Sonnet, Haiku, and a fallback, subject to compatibility.
Streaming and tool handling
The project describes support for streaming and handling tool-use or reasoning formats across compatible provider responses.
Messaging integrations
Optional Discord or Telegram workflows are described in the repository. Review setup, permissions, and security guidance before enabling them.
Local model options
Configure compatible local runtimes for workflows where you prefer to run a model on your own machine.
Open development
Browse the code, open an issue, or submit a pull request through GitHub. Contributions should follow the repository's current guidelines.
03 / Quick start
Up and running
in a few steps.
These are the package-install commands documented by the repository. Read the current README and inspect installation scripts before running commands from the internet.
Read the full quick start1
Install or update
Install from the public Git repository with uv.
uv tool install --force git+https://github.com/Alishahryar1/free-claude-code.git2
Start the server
Start the local service after completing the repository's prerequisites and configuration.
fcc-server3
Configure a provider
Follow the provider-specific instructions in the README. Keep API keys private and never commit them to Git.
Provider setup guide ↗Note: The software may be free to install, but third-party providers can charge for usage or enforce their own quotas. Do not assume that all models or API calls are free.
04 / Documentation
Need the details?
The repository README is the source of truth for installation, supported providers, configuration variables, client integrations, troubleshooting, and contribution instructions.
06 / Contact
Questions or ideas?
For website and project inquiries, email:
contact@jenkinss.spaceThis address works for incoming messages if Cloudflare Email Routing is configured correctly. Forwarding alone does not enable sending from this address.