Skip to main content

Allow HTTP requests to local AI, not just HTTPS

Feature Request

Currently, Vowen refuses to connect to local AI if the request is not https. Most, if not all, local AI installations don’t use SSL enabled http.

It would be nice to allow the use of http.

Posted by Luc Stepniewski ·
Status: Completed3 comments

Log in to comment and vote

Comments3

  • Vowen

    Team•

    Nov 29, 2025

    Hi @Luc Stepniewski We’ve added support for HTTP requests, so now calls can be made to local LLMs. It’s part of the v0.1.3 release which we just pushed, please check it out.

    Thanks again for the suggestion!

  • Vowen

    Team•

    Nov 26, 2025

    Hi @Luc Stepniewski , we’ll be adding this in our next update coming out in the next day or so (version 0.1.3). We wanted to understand how you’re using it a bit better. Are you using a local LLama setup for using AI? Wanted to understand the use case better. Thank you for the suggestion.

    • Luc Stepniewski

      •

      Nov 26, 2025

      Exactly, I’m hosting a llama-swap (for having the possibility to change the model like one would do with OpenAI) which then proxies the request to a local llama.cpp instance.