Your Own AI

Uses / The engine under your other apps

The engine under your other apps

Every AI app wants a model server; almost none give you one worth pointing at.

Your Own AI serves an OpenAI-compatible endpoint on your machine while it's open - so agent frameworks like Hermes Agent and OpenClaw, coding editors, or a ten-line script can use YOUR AIs instead of a bare model: the personality you authored answers, your memories inform it, and the conversation lands in your signed records like any other, marked as coming from that app.

Local calls need no API key - being on your machine is the credential.

Engines settings - the standard engine, one-click CUDA, and your own server

How it works

  1. 1

    Find the endpoint

    Settings → External access: the address, ready to copy, live while the app is open.

  2. 2

    Pick which AI answers

    Every AI has a model name. "Copy setup" gives the base URL, model name, and a working example for any OpenAI-compatible client.

  3. 3

    Point your app at it

    Set the base URL and model in the other app's settings - configuration, not a plugin.

  4. 4

    See the record

    Conversations from other apps appear on the Memory page with an "API" badge - the same signed records as in-app chats.

Also see

Code in your own building →Receipts for AI work →
Download free

Free · No account needed · macOS, Windows, Linux