Bug Report

Description

Bug Summary:
Image generation using DALL.E 3 on the Open WebUI interface produces significantly lower quality results compared to the ChatGPT interface with the same prompt. I feel like the response to my prompt is being used to generate the image instead of my original prompt.

Steps to Reproduce:
Open the Open WebUI interface
Configure Open AI key
Enter a prompt for image generation (for example draw me a bee on a flower).
Generate the image.
Compare the image quality with the same prompt generated via the ChatGPT interface.

Expected Behavior:
The images generated using the Open WebUI interface should match the quality and detail of those generated by the ChatGPT interface.

Actual Behavior:
The images generated by the Open WebUI interface are of much lower quality compared to those generated by the ChatGPT interface with the same prompt.

Environment

• Open WebUI Version: 0.3.5
• Operating System: Window 11
• Browser (if applicable): Chrome 120.0.6099.130

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.

Logs and Screenshots

Installation Method

Docker