[GH-ISSUE #20921] issue: think field is used incorrectly for ollama and reasoning_effort is not used correctly with its API #57994

Closed
opened 2026-05-05 22:08:16 -05:00 by GiteaMirror · 4 comments
Owner

Originally created by @svenstaro on GitHub (Jan 25, 2026).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/20921

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

v0.7.2

Ollama Version (if applicable)

v0.15.1

Operating System

Arch 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

Thinking effort should be settable for gpt-oss models when using the ollama connection.

Actual Behavior

Thinking effort is not set correctly by open-webui.

Steps to Reproduce

  1. Set ollama in open-webui
  2. Have gpt-oss-120b or gpt-oss-20b available in ollama
  3. Set the model in open-webui
  4. In the model advanced params, set Reasoning Effort to high
  5. Dispatch a request
  6. Observe ollama log, it will complain as such: msg="invalid option provided" option=reasoning_effort

Logs & Screenshots

Jan 25 10:27:27 morpheus ollama[52160]: time=2026-01-25T10:27:27.044+01:00 level=WARN source=types.go:966 msg="invalid option provided" option=reasoning_effort

Additional Information

open-webui currently incorrectly assumes that think is always bool but according to upstream docs, this can be either bool or string, depending on the model.

Observe the warning on the upstream docs:

GPT-OSS requires think to be set to "low", "medium", or "high". Passing true/false is ignored for that model.

Furthermore, open-webui will also currently try to send reasoning_effort to ollama as-is which results in this log line in ollama:

Jan 25 10:27:27 comp ollama[52160]: time=2026-01-25T10:27:27.044+01:00 level=WARN source=types.go:966 msg="invalid option provided" option=reasoning_effort
Originally created by @svenstaro on GitHub (Jan 25, 2026). Original GitHub issue: https://github.com/open-webui/open-webui/issues/20921 ### 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 v0.7.2 ### Ollama Version (if applicable) v0.15.1 ### Operating System Arch 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 Thinking effort should be settable for gpt-oss models when using the ollama connection. ### Actual Behavior Thinking effort is not set correctly by open-webui. ### Steps to Reproduce 1. Set ollama in open-webui 2. Have `gpt-oss-120b` or `gpt-oss-20b` available in ollama 3. Set the model in open-webui 4. In the model advanced params, set `Reasoning Effort` to `high` 5. Dispatch a request 6. Observe ollama log, it will complain as such: `msg="invalid option provided" option=reasoning_effort` ### Logs & Screenshots ``` Jan 25 10:27:27 morpheus ollama[52160]: time=2026-01-25T10:27:27.044+01:00 level=WARN source=types.go:966 msg="invalid option provided" option=reasoning_effort ``` ### Additional Information open-webui [currently incorrectly assumes that `think` is always `bool`](https://github.com/open-webui/open-webui/blob/2b26355002064228e9b671339f8f3fb9d1fafa73/backend/open_webui/utils/payload.py#L190) but [according to upstream docs, this can be either `bool` or `string`, depending on the model](https://docs.ollama.com/capabilities/thinking). Observe the warning on the upstream docs: > GPT-OSS requires think to be set to "low", "medium", or "high". Passing true/false is ignored for that model. Furthermore, open-webui will also currently try to send `reasoning_effort` to ollama as-is which results in this log line in ollama: ``` Jan 25 10:27:27 comp ollama[52160]: time=2026-01-25T10:27:27.044+01:00 level=WARN source=types.go:966 msg="invalid option provided" option=reasoning_effort ```
GiteaMirror added the bug label 2026-05-05 22:08:16 -05:00
Author
Owner

@owui-terminator[bot] commented on GitHub (Jan 25, 2026):

🔍 Similar Issues Found

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

  1. #20361 Issue: Large-scale model setting-related functionality fails.
    by shentong0722 • Jan 04, 2026 • bug

  2. #20327 issue: Unable to use any Open WebUI version newer than 0.6.25 due to hybrid search performance
    by galvanoid • Jan 02, 2026 • bug

  3. #19864 issue: Ollama Parameters get overriden after native tool calls
    by Haervwe • Dec 10, 2025 • bug

  4. #20231 issue: Non-streaming responses with thinking enabled are not parsed correctly (thinking content merged into final answer)
    by hamid370 • Dec 29, 2025 • bug

  5. #19755 issue: Error: 404, message='Not Found', url='http://ollama:11434/api/embed'
    by stevewillett • Dec 04, 2025 • bug

Show 5 more related issues
  1. #19017 issue: OPENAI_API_CONFIGS and OLLAMA_API_CONFIGS environment variables are not parsed from JSON
    by runixer • Nov 07, 2025 • bug

  2. #19170 issue: Settings not honored or understood
    by itguy327 • Nov 13, 2025 • bug

  3. #19752 issue: minor UI Bug: knowledge sharing
    by mahenning • Dec 04, 2025 • bug, confirmed issue

  4. #18145 issue: 0.6.33 regression
    by Ark-Levy • Oct 08, 2025 • bug

  5. #19987 issue: There is a lack of visual consistency between the home page and the chat interface.
    by i-iooi-i • Dec 16, 2025 • bug


💡 Tips:

  • If this is a duplicate, please consider closing this issue and adding any additional details to the existing one
  • If you found a solution in any of these issues, please share it here to help others

This comment was generated automatically by a bot. Please react with a 👍 if this comment was helpful, or a 👎 if it was not.

<!-- gh-comment-id:3796311307 --> @owui-terminator[bot] commented on GitHub (Jan 25, 2026): 🔍 **Similar Issues Found** I found some existing issues that might be related to this one. Please check if any of these are duplicates or contain helpful solutions: 1. [#20361](https://github.com/open-webui/open-webui/issues/20361) **Issue: Large-scale model setting-related functionality fails.** *by shentong0722 • Jan 04, 2026 • `bug`* 2. [#20327](https://github.com/open-webui/open-webui/issues/20327) **issue: Unable to use any Open WebUI version newer than 0.6.25 due to hybrid search performance** *by galvanoid • Jan 02, 2026 • `bug`* 3. [#19864](https://github.com/open-webui/open-webui/issues/19864) **issue: Ollama Parameters get overriden after native tool calls** *by Haervwe • Dec 10, 2025 • `bug`* 4. [#20231](https://github.com/open-webui/open-webui/issues/20231) **issue: Non-streaming responses with thinking enabled are not parsed correctly (thinking content merged into final answer)** *by hamid370 • Dec 29, 2025 • `bug`* 5. [#19755](https://github.com/open-webui/open-webui/issues/19755) **issue: Error: 404, message='Not Found', url='http://ollama:11434/api/embed'** *by stevewillett • Dec 04, 2025 • `bug`* <details> <summary>Show 5 more related issues</summary> 6. [#19017](https://github.com/open-webui/open-webui/issues/19017) **issue: OPENAI_API_CONFIGS and OLLAMA_API_CONFIGS environment variables are not parsed from JSON** *by runixer • Nov 07, 2025 • `bug`* 7. [#19170](https://github.com/open-webui/open-webui/issues/19170) **issue: Settings not honored or understood** *by itguy327 • Nov 13, 2025 • `bug`* 8. [#19752](https://github.com/open-webui/open-webui/issues/19752) **issue: minor UI Bug: knowledge sharing** *by mahenning • Dec 04, 2025 • `bug`, `confirmed issue`* 9. [#18145](https://github.com/open-webui/open-webui/issues/18145) **issue: 0.6.33 regression** *by Ark-Levy • Oct 08, 2025 • `bug`* 10. [#19987](https://github.com/open-webui/open-webui/issues/19987) **issue: There is a lack of visual consistency between the home page and the chat interface.** *by i-iooi-i • Dec 16, 2025 • `bug`* </details> --- 💡 **Tips:** - If this is a duplicate, please consider closing this issue and adding any additional details to the existing one - If you found a solution in any of these issues, please share it here to help others *This comment was generated automatically by a bot.* Please react with a 👍 if this comment was helpful, or a 👎 if it was not.
Author
Owner

@svenstaro commented on GitHub (Jan 25, 2026):

I think a fairly simple fix would be to map the value of the reasoning_effort field to think if reasoning_effort is set, otherwise go with the default value of think?

<!-- gh-comment-id:3796328057 --> @svenstaro commented on GitHub (Jan 25, 2026): I think a fairly simple fix would be to map the value of the `reasoning_effort` field to `think` if `reasoning_effort` is set, otherwise go with the default value of `think`?
Author
Owner

@tjbck commented on GitHub (Feb 22, 2026):

Addressed in dev!

<!-- gh-comment-id:3941888539 --> @tjbck commented on GitHub (Feb 22, 2026): Addressed in dev!
Author
Owner

@Classic298 commented on GitHub (Feb 22, 2026):

30a13b9b2f

<!-- gh-comment-id:3941890863 --> @Classic298 commented on GitHub (Feb 22, 2026): https://github.com/open-webui/open-webui/commit/30a13b9b2fb2c6da7e1ddbf52edb93a58d09cc56
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/open-webui#57994