Install and connect (LocalAI)

LocalAI is a local OpenAI-compatible API gateway over many backends. Suited to Docker users or anyone who wants one compatibility layer for multiple models.

Quick start (Docker example)

Image names and ports follow official docs; illustrative:

docker run -p 8080:8080 --name local-ai localai/localai:latest

Place or pull models per docs; confirm on the PC:

http://127.0.0.1:8080/v1/models

LAN

• After publishing the container port to the host, allow 8080 (or your mapped port) in the firewall.
• On the phone: http://YOUR_PC_IP:8080/v1.
• If the container binds the wrong interface, check Docker port publish and host firewall.

For MiniTavern

Field	Value
Model platform	Custom (OpenAI)
Base URL	http://YOUR_PC_IP:8080/v1
API Key	Per whether LocalAI auth is enabled
Model name	Same model id as in LocalAI config

Next

Configure in MiniTavern · Tools overview · Troubleshoot