I’ve been using Open WebUI as my main frontend for interacting with large language models. It’s open-source, flexible, and future-proof enough to accommodate new models and use cases as they appear. Recently I started exploring Agno, an open-source agent framework with solid documentation and lots of examples. Here’s a quick, repeatable setup to run Agno agents from Open WebUI.

  • Open WebUI installed and running
  • Internet access to fetch the Agno integration script
  • A basic familiarity with Open WebUI’s UI (import/install scripts)
  • An Agno API instance running (see the examples for a quick start)

I published the script on the official Open WebUI Homepage: Link.

  1. Obtain the Agno integration script
    1. Get the script from the link above.
    2. Import the script into Open WebUI.
  2. In Open WebUI, open Administration panel and choose Functions.
    1. Provide the path/URL to the Agno instance including path and confirm.
    2. Configure the integration, see below.
  3. Save

  • Agno Agents listed next to your LLM models
  • A responsive chat interface with Agno-based behavior

  • Agent not appearing: re-import the script, then restart the UI.
  • Agent not responding: check the network access for the script, verify model selection, and inspect any error logs in the UI.
  • Mismatched versions: ensure your Open WebUI version is compatible with the Agno script version.

Happy chatting with your Agno Agents!