I built Infercat: Share your local AI with friends over an encrypted p2p tunnel. Connect with an invite code from anywhere.

Infercat allows users to share their local AI models, such as those running on llama.cpp, vLLM, Ollama, and LM Studio, with friends over an encrypted peer-to-peer tunnel. It provides automatic or configurable discovery and generates an OpenAI-compatible gateway at http://127.0.0.1:11435/v1 for tools like Open WebUI, Cursor, Claude Code, or curl. Infercat is open-source under the MIT License.

Unlike other local AI sharing methods, Infercat offers an encrypted P2P tunnel and an OpenAI-compatible gateway, simplifying access for various tools.