Fix tool search failing with reasoning models#44
Merged
Conversation
Reasoning models (o3, o4-mini, gpt-5*) reject the max_tokens parameter in the Chat Completions API and require max_completion_tokens instead. When the user's active model is a reasoning model, the tool search service inherits that model and the API call fails with a 400 error, preventing tool discovery entirely. Tool search is a lightweight classification task that doesn't benefit from reasoning. For OpenAI reasoning models, automatically downgrade to gpt-4.1-mini which is cheaper, faster, and uses the standard max_tokens parameter without issues.
vl3c
force-pushed
the
fix/tool-search-reasoning-model-compat
branch
from
3c50ce3 to
331d70e
Compare
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
max_tokens is not supported with this modelmax_completion_tokensbudget — 500 tokens may not leave enough room for the actual JSON outputgpt-4.1-minifor tool search — it's a lightweight classification task that doesn't benefit from reasoning, and this avoids both the parameter incompatibility and the token budget issueTest plan
static/tool_search_service.pyshear_object🤖 Generated with Claude Code