Infinity Download

Setup guide · Any server

Any OpenAI-compatible server.

Jan, llama.cpp, vLLM, a server at work, or any service that speaks the OpenAI API.

What you need

The server’s base URL (it usually ends in /v1) and, if it needs one, a key.

Set it up

  1. Open Settings ▸ Accounts and models and choose Custom (OpenAI-compatible)….
  2. Paste the server’s Base URL, for example http://localhost:8080/v1. Add a key if the server needs one.
  3. Press Verify. Infinity lists the server’s models; pick one and chat.

Good to know

  • Jan, llama.cpp and GPT4All are found automatically when they’re running. In Jan, turn on Settings ▸ Local API Server.
  • A server on your own Mac or network keeps your chats there. A cloud service gets what you send it, under its own terms.
Stuck?See common fixes, or tell us what happened.

Other guides

Ollama · LM Studio · Apple Intelligence · MLX LM · Your ChatGPT plan · Your Claude plan · OpenAI · Anthropic · Google Gemini · OpenRouter · Hugging Face · Groq · Mistral · xAI · Together AI