[GH-ISSUE #24830] issue: Workspace Model overrides default/main model settings #107402

Open
opened 2026-05-18 06:12:37 -05:00 by GiteaMirror · 1 comment
Owner

Originally created by @frenzybiscuit on GitHub (May 17, 2026).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/24830

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

0.9.5

Ollama Version (if applicable)

No response

Operating System

Linux

Browser (if applicable)

No response

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

Default settings / Main Model should take precedence over custom model settings

Actual Behavior

Custom model settings override Default Settings / Main Model

Steps to Reproduce

Set default parameters to 256 max_tokens
Set model parameters to 256 max_tokens
Create a custom model under Workspace and set it to 2048 max_tokens
User can still modify max_tokens to any number on the custom model
The backend receives the modified max_tokens from the end user, and is not limited by the default settings & main model.

Logs & Screenshots

Backend:

Image

Additional Information

This is a problem since some of my users are setting max_tokens to 10k+

Originally created by @frenzybiscuit on GitHub (May 17, 2026). Original GitHub issue: https://github.com/open-webui/open-webui/issues/24830 ### Check Existing Issues - [x] I have searched for any existing and/or related issues. - [x] I have searched for any existing and/or related discussions. - [x] 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!). - [x] I am using the latest version of Open WebUI. ### Installation Method Docker ### Open WebUI Version 0.9.5 ### Ollama Version (if applicable) _No response_ ### Operating System Linux ### Browser (if applicable) _No response_ ### Confirmation - [x] I have read and followed all instructions in `README.md`. - [x] I am using the latest version of **both** Open WebUI and Ollama. - [x] I have included the browser console logs. - [x] I have included the Docker container logs. - [x] I have **provided every relevant configuration, setting, and environment variable used in my setup.** - [x] 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). - [x] 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 Default settings / Main Model should take precedence over custom model settings ### Actual Behavior Custom model settings override Default Settings / Main Model ### Steps to Reproduce Set default parameters to 256 max_tokens Set model parameters to 256 max_tokens Create a custom model under Workspace and set it to 2048 max_tokens User can still modify max_tokens to any number on the custom model The backend receives the modified max_tokens from the end user, and is not limited by the default settings & main model. ### Logs & Screenshots Backend: <img width="488" height="438" alt="Image" src="https://github.com/user-attachments/assets/7ff5e8e8-57ae-416d-8c71-25524c7380a2" /> ### Additional Information This is a problem since some of my users are setting max_tokens to 10k+
GiteaMirror added the bug label 2026-05-18 06:12:37 -05:00
Author
Owner

@owui-terminator[bot] commented on GitHub (May 17, 2026):

🔍 Related Issues Found

I found some existing issues that might be related. Please check if any of these are duplicates or contain helpful solutions:

  1. 🟣 #19214 issue: default model params take precedence over user-specified model params
    This is the closest match: it reports the same precedence bug where default/model admin parameters override user-specified values. The new issue is about default/main model settings overriding custom workspace model settings for max_tokens, which is the same kind of parameter-precedence problem.
    by Simon-Stone · bug

  2. 🟣 #17151 issue: user setting overrules model settings for websearch
    This issue shows the same general override pattern for model-specific restrictions being bypassed by a global setting. While it concerns websearch rather than token limits, it is related because both describe workspace/model configuration being overridden by higher-level defaults in an unexpected way.
    by Maximilian-Pichler · bug

  3. 🟣 #18619 issue: Nested options.max_tokens in model parameters not recursively converted to num_predict
    This is related because it involves max_tokens in workspace/model parameters and how those settings are transformed before being sent downstream. It may not be the exact same precedence bug, but it is directly adjacent to the max_tokens handling involved in the new report.
    by elazar · bug

  4. 🟣 #13149 issue: 400: Invalid 'stop': array too long when using Workspace model specifically
    This issue is workspace-model-specific and shows that model-level configuration can behave differently when a model is wrapped in Workspace. Even though the symptom is about stop arrays, it is relevant as another workspace-model configuration bug affecting request payloads.
    by theprezz87 · bug


💡 If your issue is a duplicate, please close it and add any additional details to the existing issue instead.

This comment was generated automatically. React with 👍 if helpful, 👎 if not.

<!-- gh-comment-id:4470160359 --> @owui-terminator[bot] commented on GitHub (May 17, 2026): <!-- terminator-bot:related-issues-reply --> 🔍 **Related Issues Found** I found some existing issues that might be related. Please check if any of these are duplicates or contain helpful solutions: 1. 🟣 [#19214](https://github.com/open-webui/open-webui/issues/19214) **issue: default model params take precedence over user-specified model params** *This is the closest match: it reports the same precedence bug where default/model admin parameters override user-specified values. The new issue is about default/main model settings overriding custom workspace model settings for `max_tokens`, which is the same kind of parameter-precedence problem.* *by Simon-Stone · `bug`* 2. 🟣 [#17151](https://github.com/open-webui/open-webui/issues/17151) **issue: user setting overrules model settings for websearch** *This issue shows the same general override pattern for model-specific restrictions being bypassed by a global setting. While it concerns websearch rather than token limits, it is related because both describe workspace/model configuration being overridden by higher-level defaults in an unexpected way.* *by Maximilian-Pichler · `bug`* 3. 🟣 [#18619](https://github.com/open-webui/open-webui/issues/18619) **issue: Nested `options.max_tokens` in model parameters not recursively converted to `num_predict`** *This is related because it involves `max_tokens` in workspace/model parameters and how those settings are transformed before being sent downstream. It may not be the exact same precedence bug, but it is directly adjacent to the `max_tokens` handling involved in the new report.* *by elazar · `bug`* 4. 🟣 [#13149](https://github.com/open-webui/open-webui/issues/13149) **issue: 400: Invalid 'stop': array too long when using Workspace model specifically** *This issue is workspace-model-specific and shows that model-level configuration can behave differently when a model is wrapped in Workspace. Even though the symptom is about `stop` arrays, it is relevant as another workspace-model configuration bug affecting request payloads.* *by theprezz87 · `bug`* --- 💡 If your issue is a duplicate, please close it and add any additional details to the existing issue instead. *This comment was generated automatically.* React with 👍 if helpful, 👎 if not.
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/open-webui#107402