token.place
A secure, peer-to-peer generative AI platform that matches people who want to run LLM inference with volunteers donating their spare compute.
Visit website (opens in a new tab)token.place
token.place is built around a relay component and a compute-node component; the relay can be run alone for a fast local path, or the full stack can be run for development with tests and pre-commit checks.
It is written in Python with a Node.js-based tooling layer, and its continuous integration covers linting, tests, documentation and coverage, plus dependency, code-scanning and secret-scanning checks.
It installs from source with a Python virtual environment and npm, and is MIT licensed.