[GH-ISSUE #24745] issue: Incorrect model used for title auto-generation when configured to Current Model, and the current model is changed in the dropdown menu #91130

Open
opened 2026-05-15 16:24:35 -05:00 by GiteaMirror · 3 comments
Owner

Originally created by @rotemdan on GitHub (May 15, 2026).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/24745

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

Latest dev version

Ollama Version (if applicable)

No response

Operating System

Windows 11

Browser (if applicable)

Latest Brave

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 title auto-generation model is set to Current Model, it should use the model selected in the main model dropdown to generate the title, in any case, including in the case where the current prompt was regenerated or edited several times with different models.

Actual Behavior

What currently happens is that clicking Rename and then the stars icon to regenerate the title, the model selected for prompt regeneration is an arbitrary model somewhere from the history of models use for previous revisions of the prompt(s) for the current thread (I'm not sure but I think likely the first one used in the first revision), not the currently selected one.

Steps to Reproduce

  1. Create a new thread
  2. Prompt several times with multiple different models
  3. Click the Rename button and the stars icon to regenerate the title, and observe which model is used for the task
  4. Repeat by changing the active model using the dropdown several times and keep observing which model is used to regenerate the title

You should see that the model selected for title auto-generation is not impacted by the current model selected in the dropdown list, as well as the last model used to regenerate the response to the prompt.

Logs & Screenshots

(No relevant error messages or logs)

Additional Information

I initially reported the behavior in later messages in #24604, which was marked as fixed, but did not resolve the issue.

Due to the framing of that issue around a particular different error (which may or may not have been fixed), and being closed, I decided to open a new issue dedicated to this particular aspect of the problem, with the correct framing in the issue title.

Originally created by @rotemdan on GitHub (May 15, 2026). Original GitHub issue: https://github.com/open-webui/open-webui/issues/24745 ### 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 Latest dev version ### Ollama Version (if applicable) _No response_ ### Operating System Windows 11 ### Browser (if applicable) Latest Brave ### 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 When title auto-generation model is set to `Current Model`, it should use the model selected in the main model dropdown to generate the title, in any case, including in the case where the current prompt was regenerated or edited several times with different models. ### Actual Behavior What currently happens is that clicking `Rename` and then the stars icon to regenerate the title, the model selected for prompt regeneration is an arbitrary model somewhere from the history of models use for previous revisions of the prompt(s) for the current thread (I'm not sure but I think likely the first one used in the first revision), not the currently selected one. ### Steps to Reproduce 1. Create a new thread 2. Prompt several times with multiple different models 3. Click the `Rename` button and the stars icon to regenerate the title, and observe which model is used for the task 4. Repeat by changing the active model using the dropdown several times and keep observing which model is used to regenerate the title You should see that the model selected for title auto-generation is not impacted by the current model selected in the dropdown list, as well as the last model used to regenerate the response to the prompt. ### Logs & Screenshots (No relevant error messages or logs) ### Additional Information I initially reported the behavior in later messages in #24604, which was marked as fixed, but did not resolve the issue. Due to the framing of that issue around a particular different error (which may or may not have been fixed), and being closed, I decided to open a new issue dedicated to this particular aspect of the problem, with the correct framing in the issue title.
GiteaMirror added the bug label 2026-05-15 16:24:35 -05:00
Author
Owner

@owui-terminator[bot] commented on GitHub (May 15, 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. 🟣 #24604 issue: Error "Model '' was not found" when auto-title generation is invoked
    This is the direct precursor issue the reporter references. It describes auto-title generation using the wrong model history and failing when the selected model is unavailable, which is closely related to the same title-generation model selection path.
    by rotemdan · bug

  2. 🟢 #14681 issue: Current model set as Task model is reloaded to generate title with different keep alive parameter
    This open bug is about the title-generation task reloading/using the current model after chat generation, showing the same 'current model' task-model flow as this issue. It appears to cover the same subsystem and could be a related regression in how the task model is chosen.
    by trinhkvo · bug

  3. 🟣 #16768 Issue #16768
    This closed issue explicitly concerns title/tag generation offloading the current model, which is the same feature area. It likely overlaps with bugs in how title generation picks the active/current model.
    by unknown

  4. 🟣 #13055 Issue #13055
    This feature request is about making the API model used for title generation configurable, so it touches the same title-generation model-selection mechanism. It may provide historical context for why the current-model behavior exists and where it can break.
    by unknown


💡 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:4456598851 --> @owui-terminator[bot] commented on GitHub (May 15, 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. 🟣 [#24604](https://github.com/open-webui/open-webui/issues/24604) **issue: Error "Model '' was not found" when auto-title generation is invoked** *This is the direct precursor issue the reporter references. It describes auto-title generation using the wrong model history and failing when the selected model is unavailable, which is closely related to the same title-generation model selection path.* *by rotemdan · `bug`* 2. 🟢 [#14681](https://github.com/open-webui/open-webui/issues/14681) **issue: Current model set as Task model is reloaded to generate title with different keep alive parameter** *This open bug is about the title-generation task reloading/using the current model after chat generation, showing the same 'current model' task-model flow as this issue. It appears to cover the same subsystem and could be a related regression in how the task model is chosen.* *by trinhkvo · `bug`* 3. 🟣 [#16768](https://github.com/open-webui/open-webui/issues/16768) **Issue #16768** *This closed issue explicitly concerns title/tag generation offloading the current model, which is the same feature area. It likely overlaps with bugs in how title generation picks the active/current model.* *by unknown* 4. 🟣 [#13055](https://github.com/open-webui/open-webui/issues/13055) **Issue #13055** *This feature request is about making the API model used for title generation configurable, so it touches the same title-generation model-selection mechanism. It may provide historical context for why the current-model behavior exists and where it can break.* *by unknown* --- 💡 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.
Author
Owner

@Classic298 commented on GitHub (May 15, 2026):

investigating

<!-- gh-comment-id:4463181174 --> @Classic298 commented on GitHub (May 15, 2026): investigating
Author
Owner

@Classic298 commented on GitHub (May 15, 2026):

https://github.com/open-webui/open-webui/pull/24796

<!-- gh-comment-id:4463471707 --> @Classic298 commented on GitHub (May 15, 2026): https://github.com/open-webui/open-webui/pull/24796
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/open-webui#91130