[GH-ISSUE #18863] feat: payload size limitations for the API #18697

Closed
opened 2026-04-20 00:54:31 -05:00 by GiteaMirror · 0 comments
Owner

Originally created by @JIW-roosi on GitHub (Nov 2, 2025).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/18863

Check Existing Issues

  • I have searched all existing open AND closed issues and discussions for similar requests. I have found none that is comparable to my request.

Problem Description

A user was using the open webui API for code generation and refactoring in SQL Server Management Studio.
The prompts injected were massive (roughly 1 million characters from a tabular/semantic model) and some queries took more than 60 hours to complete.
This was not noticed initially, as the user was experimenting on a Friday and effects were only seen on Monday; loading was extremely slow when it worked but was unresponsive most of the time.
Container memory use was continuously higher than 7GB, even going north of 10GB without any other user being able to get a response, when it normally is below 2GB.
A force-logout and jwt token reset was necessary to fix the situation.

Desired Solution you'd like

payload size limitations for the API, as optimisation of large request handling to such an extent seems unfeisable

Alternatives Considered

No response

Additional Context

open webui v0.6.32 in a k8s deployment
endpoint used: /api/chat/completions, auth through jwt token

Originally created by @JIW-roosi on GitHub (Nov 2, 2025). Original GitHub issue: https://github.com/open-webui/open-webui/issues/18863 ### Check Existing Issues - [x] I have searched all existing open AND closed issues and discussions for similar requests. I have found none that is comparable to my request. ### Problem Description A user was using the open webui API for code generation and refactoring in SQL Server Management Studio. The prompts injected were massive (roughly 1 million characters from a tabular/semantic model) and some queries took more than 60 hours to complete. This was not noticed initially, as the user was experimenting on a Friday and effects were only seen on Monday; loading was extremely slow when it worked but was unresponsive most of the time. Container memory use was continuously higher than 7GB, even going north of 10GB without any other user being able to get a response, when it normally is below 2GB. A force-logout and jwt token reset was necessary to fix the situation. ### Desired Solution you'd like payload size limitations for the API, as optimisation of large request handling to such an extent seems unfeisable ### Alternatives Considered _No response_ ### Additional Context open webui v0.6.32 in a k8s deployment endpoint used: /api/chat/completions, auth through jwt token
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/open-webui#18697