Check Existing Issues

• I have searched for any existing and/or related issues.
• I have searched for any existing and/or related discussions.
• I have also searched in the CLOSED issues AND CLOSED discussions and found no related items (your issue might already be addressed on the development branch!).
• I am using the latest version of Open WebUI.

Installation Method

Docker

Open WebUI Version

v0.8.2

Ollama Version (if applicable)

No response

Operating System

Ubuntu 24.04.4 LTS

Browser (if applicable)

chrome

Confirmation

• I have read and followed all instructions in README.md.
• I am using the latest version of both Open WebUI and Ollama.
• I have included the browser console logs.
• I have included the Docker container logs.
• I have provided every relevant configuration, setting, and environment variable used in my setup.
• I have clearly listed every relevant configuration, custom setting, environment variable, and command-line option that influences my setup (such as Docker Compose overrides, .env values, browser settings, authentication configurations, etc).
• I have documented step-by-step reproduction instructions that are precise, sequential, and leave nothing to interpretation. My steps:
• Start with the initial platform/version/OS and dependencies used,
• Specify exact install/launch/configure commands,
• List URLs visited, user input (incl. example values/emails/passwords if needed),
• Describe all options and toggles enabled or changed,
• Include any files or environmental changes,
• Identify the expected and actual result at each stage,
• Ensure any reasonably skilled user can follow and hit the same issue.

Expected Behavior

When attaching a PNG/JPG in chat and prompting the model, the request to the model should include the image (vision input) and the model should respond based on the image.

Actual Behavior

The model behaves as if no image was provided.
In browser DevTools, the request payload to /api/chat/completions contains features.image: false and does not include any image content parts.

Steps to Reproduce

1. Start a new chat.
2. Select model gpt-5-mini (vision enabled in model settings).
3. Attach a .png or .jpg image in the message and send a prompt like: "Please describe the image".
4. Observe: assistant responds as if no image was attached.
5. Open Browser DevTools → Network → POST /api/chat/completions → Payload:
   • features.image: false
   • messages[].content contains only type: "text" (no image parts like image_url / input_image).

Logs & Screenshots

Image uploads are stored with:

• POST /api/v1/files/?process=false
  and logs show image/png False / image/jpeg False.

Example:
... file.content_type: image/jpeg False
... "POST /api/v1/files/?process=false HTTP/1.1" 200
... file.content_type: image/png False
... "POST /api/v1/files/?process=false HTTP/1.1" 200

ki.digatron.de-1771251643888.log

Evidence (DevTools Payload)

Request: POST /api/chat/completions

• features.image: false
• no image parts present in messages

root@srv-gpt:/opt/ai# docker logs -f --since 1m ai-open-webui-ki 2>&1
| grep -nEi ' /api/v1/files/|\bupload\b|process/status|process=true|process=false|file.content_type|content HTTP'
16:2026-02-16 14:14:07.220 | INFO | open_webui.routers.files:upload_file_handler:230 - file.content_type: image/png False
17:2026-02-16 14:14:07.238 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 178.157.47.45:0 - "POST /api/v1/files/?process=false HTTP/1.1" 200
18:2026-02-16 14:14:07.299 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 178.157.47.45:0 - "GET /api/v1/files/af816a38-dba8-4e6b-a67b-c61b1026d6c7/process/status?stream=true HTTP/1.1" 200
19:2026-02-16 14:14:07.365 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 178.157.47.45:0 - "GET /api/v1/files/af816a38-dba8-4e6b-a67b-c61b1026d6c7/content HTTP/1.1" 200

Additional Information

• Worked on v0.7.2, broken since upgrade to v0.8.2
• Same setup / same browser / same model provider
• Still works with pdf , csv or docx