You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Many users query Claude and Gemini models through openai-compatible apis, while current "thinking" controllers for Claude and Gemini do not allow users to change the parameters like thinking_budget.
The reasoning_effort parameter user input in the UI is not used actually.
Implementation
Rewrite the thinking budget UI and parameters utilization.
Use the reasoning_effort parameter from the UI if exists.
Screenshots
How to Test
Use claude models, e.g., claude-sonnet-4 and test them with different thinking_budget, where 0 means no thinking.
The improvement for Gemini is hard to test, maybe different reasoning_effort parameters result different token usages.
…viders
- Add ThinkingBudgetToggle component with configurable token limits
- Implement thinking configuration in OpenAI API requests
- Update model info to support thinking mode and max thinking tokens
- Add internationalization support across 18 languages
- Ensure compatibility with Claude's thinking budget feature
Merging this PR will not cause a version bump for any packages. If these changes should not result in a new version, you're good to go. If these changes should result in a version bump, you need to add a changeset.
This PR includes no changesets
When changesets are added to this PR, you'll see the packages that this PR includes changesets for and the associated semver types
Hi @xihuai18! Thanks for this! It does seem that this breaks the reasoning sliders when using claude through the kilocode/openrouter/anthropic provider. These options seem removed in this PR, was that intentional? We do need a way to set max tokens in those providers as well:
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Context
thinking_budget.Implementation
Screenshots
How to Test