I have searched the existing issues and discussions.
Problem Description
Currently, users cannot easily toggle deep thinking modes during conversations. Different models implement this feature differently: some use boolean parameters (e.g., enable_thinking), others require special tags like [thinking], and some support multiple modes (disabled/enabled/auto). Manually adjusting these per model disrupts workflow efficiency.
Desired Solution you'd like
Add a dedicated Deep Thinking toggle in the chat input toolbar with three options:
Disabled (default)
Enabled (forces deep thinking)
Auto (model decides when to activate)
The toggle should dynamically adapt to backend models by:
Sending enable_thinking parameters for API-based models
Injecting [thinking] tags for tag-dependent models
Mapping to model-specific implementations (e.g., thinking.type=auto)
Alternatives Considered
Global settings: Less flexible for per-message control.
Manual parameter editing: Too technical for average users.
Model-specific UIs: Fragments user experience.
Originally created by @MCQSJ on GitHub (Sep 6, 2025).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/17239
### Check Existing Issues
- [x] I have searched the existing issues and discussions.
### Problem Description
Currently, users cannot easily toggle deep thinking modes during conversations. Different models implement this feature differently: some use boolean parameters (e.g., enable_thinking), others require special tags like [thinking], and some support multiple modes (disabled/enabled/auto). Manually adjusting these per model disrupts workflow efficiency.
### Desired Solution you'd like
Add a dedicated Deep Thinking toggle in the chat input toolbar with three options:
Disabled (default)
Enabled (forces deep thinking)
Auto (model decides when to activate)
The toggle should dynamically adapt to backend models by:
Sending enable_thinking parameters for API-based models
Injecting [thinking] tags for tag-dependent models
Mapping to model-specific implementations (e.g., thinking.type=auto)
### Alternatives Considered
Global settings: Less flexible for per-message control.
Manual parameter editing: Too technical for average users.
Model-specific UIs: Fragments user experience.
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
Originally created by @MCQSJ on GitHub (Sep 6, 2025).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/17239
Check Existing Issues
Problem Description
Currently, users cannot easily toggle deep thinking modes during conversations. Different models implement this feature differently: some use boolean parameters (e.g., enable_thinking), others require special tags like [thinking], and some support multiple modes (disabled/enabled/auto). Manually adjusting these per model disrupts workflow efficiency.
Desired Solution you'd like
Add a dedicated Deep Thinking toggle in the chat input toolbar with three options:
Disabled (default)
Enabled (forces deep thinking)
Auto (model decides when to activate)
The toggle should dynamically adapt to backend models by:
Sending enable_thinking parameters for API-based models
Injecting [thinking] tags for tag-dependent models
Mapping to model-specific implementations (e.g., thinking.type=auto)
Alternatives Considered
Global settings: Less flexible for per-message control.
Manual parameter editing: Too technical for average users.
Model-specific UIs: Fragments user experience.
@tjbck commented on GitHub (Sep 6, 2025):
This should be implemented as Functions.
@Classic298 commented on GitHub (Sep 6, 2025):
@MCQSJ
https://docs.openwebui.com/features/plugin/functions/filter