[GH-ISSUE #24175] issue: Model access issue with private base models (works with older models, fails with newer ones) #58887

Closed
opened 2026-05-06 00:21:10 -05:00 by GiteaMirror · 1 comment
Owner

Originally created by @Teimetues on GitHub (Apr 27, 2026).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/24175

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

v0.9.2

Ollama Version (if applicable)

No response

Operating System

Ubuntu 24.04.4

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

The assistant should work regardless of whether the underlying base model is private, as long as the user has access to the assistant itself

Actual Behavior

Newer models (e.g. GPT-5.5) seem to require the base model to be public, otherwise they are not accessible through the assistant

Steps to Reproduce

  1. Add a base model (e.g. GPT-5.5)
  2. Set the base model visibility to private
  3. Create a custom assistant (wrapper model) that uses this base model
  4. Log in as a non-admin user
  5. Try to use the assistant

Logs & Screenshots

Image

2026-04-27 11:43:55.132 | WARNING | open_webui.main:chat_completion:1839 - Error processing chat metadata: Model not found

Additional Information

Additional Information:

  • This issue only occurs with newer models (e.g. GPT-5.5), older models (e.g. GPT-5.4) work as expected
  • The problem seems related to access control / permissions of the base model
  • It appears that the assistant now requires access to the underlying base model, even if the user only interacts with the assistant
  • This behavior is inconsistent with previous versions

Possible cause:

  • Recent changes in model access control or RBAC may require explicit access to the base model

This might be a bug or an unintended side effect of recent permission changes.

Originally created by @Teimetues on GitHub (Apr 27, 2026). Original GitHub issue: https://github.com/open-webui/open-webui/issues/24175 ### 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 v0.9.2 ### Ollama Version (if applicable) _No response_ ### Operating System Ubuntu 24.04.4 ### 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 The assistant should work regardless of whether the underlying base model is private, as long as the user has access to the assistant itself ### Actual Behavior Newer models (e.g. GPT-5.5) seem to require the base model to be public, otherwise they are not accessible through the assistant ### Steps to Reproduce 1. Add a base model (e.g. GPT-5.5) 2. Set the base model visibility to private 3. Create a custom assistant (wrapper model) that uses this base model 4. Log in as a non-admin user 5. Try to use the assistant ### Logs & Screenshots <img width="1061" height="198" alt="Image" src="https://github.com/user-attachments/assets/284d1d4c-92fd-4e53-b812-ccaee12d408d" /> 2026-04-27 11:43:55.132 | WARNING | open_webui.main:chat_completion:1839 - Error processing chat metadata: Model not found ### Additional Information **Additional Information:** * This issue only occurs with newer models (e.g. GPT-5.5), older models (e.g. GPT-5.4) work as expected * The problem seems related to access control / permissions of the base model * It appears that the assistant now requires access to the underlying base model, even if the user only interacts with the assistant * This behavior is inconsistent with previous versions Possible cause: * Recent changes in model access control or RBAC may require explicit access to the base model This might be a bug or an unintended side effect of recent permission changes.
GiteaMirror added the bug label 2026-05-06 00:21:10 -05:00
Author
Owner

@Teimetues commented on GitHub (Apr 27, 2026):

"Use the search this was reported 3 more times. It's a security bug that was fixed. If you don't have access to the base model you shouldn't have access to it though the workspace model either

Just turn the base models to public but hide them (three dots > hide model)"

<!-- gh-comment-id:4327176303 --> @Teimetues commented on GitHub (Apr 27, 2026): "Use the search this was reported 3 more times. It's a security bug that was fixed. If you don't have access to the base model you shouldn't have access to it though the workspace model either Just turn the base models to public but hide them (three dots > hide model)"
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/open-webui#58887