Skip to main content

Custom LLM provider support

Feature RequestsHigh PriorityAcknowledged

Add the option to change the LLM API provider, mainly to be able to plug in a Ollama instance (or other OpenAI compatible API).

With this, we’ll be able to have a fully local and private Omi workflow

Posted by An Anonymous User ·
Status: Under Consideration2 comments

Log in to comment and vote

Comments2

  • Aryan

    Team•

    Oct 4

    Thanks for clarifying the privacy goal. You’re asking to point Omi’s own AI features at an Ollama instance or another OpenAI-compatible endpoint not to export conversations to a separate automation. Our earlier suggestion doesn’t provide that experience.

    We’re keeping custom LLM provider support under consideration, with no release date to share. Which parts would you need to run locally chat, summaries, memory processing, transcription, or all of them? And are you using Omi on iOS, Android, or desktop? That will help define the request accurately.

  • L1 Support Team - Omi AI

    •

    Dec 26, 2025

    That's a great idea. You can have a similar experience by using automations to pull your conversations via API and send them to the LLM you want. See https://docs.omi.me/doc/developer/api/overview

    You can also use your own transcription service, set it up at Settings > Dev Settings > Transcription > Bring your own.

    Passing this to our team so they can think about it. Please keep giving us your valuable feedback!