FreeLLMAPI
A local server that spreads requests across many free LLM providers behind one OpenAI-style endpoint.
- Stars
- 46k
- Forks
- 6.3k
- Licence
- MIT
- Language
- TypeScript
- Last update
- 4 days ago
- Type
- Dev tool
What it is
FreeLLMAPI is a program you run on your own machine or server. You add API keys from several providers that offer free tiers, and it exposes a single OpenAI-compatible /v1 endpoint. A router picks an available model for each request, switches to another provider when one is rate-limited, and tracks usage so you stay within free-tier caps. Keys are stored encrypted. The README describes it as intended for personal experimentation.
- Send requests from an OpenAI client library to a local server and let it choose a provider
- Configure a coding agent such as Claude Code or Aider to use the unified endpoint
- Add a custom OpenAI-compatible endpoint, such as a local Ollama or LM Studio server, alongside the free providers
- Track per-key usage in the dashboard to stay under free-tier limits
Great for
- Developers who want to experiment with several free LLM providers through one API
- People pointing OpenAI-compatible tools or coding agents at a single local endpoint
- Students or hobbyists testing models without paying for usage
- Anyone who wants automatic failover when one free provider hits a rate limit
Not ideal for
- Production services that need reliable uptime, since the README frames it as personal experimentation only
- Users who need guaranteed quotas or SLAs, because free tiers can change without notice
- People who do not want to manage API keys and a local server
How to get it
- 1
Run with Docker
You'll need: Docker (for the self-host option shown in the README)
Clone the repository
git clone https://github.com/tashfeenahmed/freellmapi - 2
Run it
- Install the app or start it with Docker using the instructions in the install guide
- Open the local dashboard and add API keys for the providers you have accounts with
- Point an OpenAI-compatible client at the local /v1 endpoint using your unified key
- Send a test request and check the usage view to confirm it routed correctly
- 3
Get the most out of it
- Add keys for several providers so failover has somewhere to go when one is rate-limited
- Use the setup command for supported agents, such as npx freellmapi setup-claude, to configure them automatically
- Add a custom OpenAI-compatible endpoint for local models from the Keys page
- Read the API reference and client guides in the docs folder for per-tool recipes
Downloads
Strengths
- Gives one OpenAI-compatible endpoint for many providers, so existing client code needs little change
- Automatic failover and routing reduce the effort of handling per-provider rate limits
- Keys are stored encrypted and usage is tracked per key
Trade-offs
- The README says it is for personal experimentation only, not production use
- Free tier availability and quotas are controlled by third-party providers and can change
- Some features, such as the live premium catalog, are paid ($19/yr per the README)
How it compares
| Project | Stars | Type | Level | Updated | Best for |
|---|---|---|---|---|---|
| freellmapi (this) | 46k | Dev tool | intermediate | 4 days ago | Developers who want to experiment with several free LLM providers through one API |
| ollama | 186k | App | beginner | 4 days ago | Developers who want a local model API for their own apps |
Stats synced from GitHub. The assessment above was drafted with AI from the project's README on 2026-10-29 and may miss recent changes — check the README before relying on a command. MIT licence: free to use, modify and distribute, including commercially, as long as you keep the copyright notice.