Bug Report

Various vision-capable models (Llama3, phi3) I've tried ignore the images parameter of the API call /chat. They say things like:

As mentioned earlier, without an actual visual provided for "[img-0]" or "this one," it's impossible for me to give a description. However, if these are placeholders for images that should have been included with your message, please make sure the images are attached properly when you resubmit your query. Once they are visible, I can certainly assist in interpreting and describing them!

It was working first, and now it's no longer working.

No relevant log messages from ollama or Open-WebUI.

Description

Bug Summary:

Chat messages with uploaded images are not recognized as having images.

Steps to Reproduce:

Select image (phone / desktop doesn't matter), type question about image, get error above.

Expected Behavior:

Get image description as I used to get.

Actual Behavior:

See above.

Environment

• Open WebUI Version: ghcr.io/open-webui/open-webui:main at the time of this bug report

• Ollama (if applicable): 0.1.44 (installed today)

• Operating System: Lunix (Fedora 40)

• Browser (if applicable): Chromium on mobile, Firefox on desktop.

Reproduction Details

Confirmation:

• I have read and followed all the instructions provided in the README.md.
• I am on the latest version of both Open WebUI and Ollama.
• I have included the browser console logs.
• I have included the Docker container logs.

Logs and Screenshots

Browser Console Logs:
[Include relevant browser console logs, if applicable]

Docker Container Logs:
[Include relevant Docker container logs, if applicable]

Screenshots (if applicable):
[Attach any relevant screenshots to help illustrate the issue]

Installation Method

Docker.

Additional Information

None.

Note

If the bug report is incomplete or does not follow the provided instructions, it may not be addressed. Please ensure that you have followed the steps outlined in the README.md and troubleshooting.md documents, and provide all necessary information for us to reproduce and address the issue. Thank you!