fix: Google Gemini thinking tokens now included in cost calculation #336
Conversation
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
Fixes two issues with Google Gemini on Vertex AI: 1. Typo fix: `thinkingTokenCount` → `thoughtsTokenCount` in gemini.ex The field name was wrong so reasoning tokens were never extracted. 2. Gemini thinking tokens weren't being costed because they're reported separately from output tokens (candidatesTokenCount excludes thoughtsTokenCount). This fix is Google Gemini-specific only. All other providers maintain their current behavior unchanged via a conservative default (add_reasoning_to_cost: false).
The original implementation had a bug where is_google_gemini_format() checked for 'thoughtsTokenCount' key, but providers already normalize this to 'reasoning_tokens' before normalize_usage() runs. This fix: - Has Google/Vertex Gemini providers set add_reasoning_to_cost: true explicitly in their extract_usage/2 callbacks - Updates normalize_usage() to preserve provider-set flags via get_add_reasoning_to_cost() helper - Adds comprehensive tests for the usage normalization pipeline
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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Outdated suggestions cannot be applied.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
gemini.ex:thinkingTokenCount→thoughtsTokenCount(field name was wrong so reasoning tokens were never extracted)candidatesTokenCount)Scope
This fix is Google Gemini-specific only. All other providers (OpenAI, Azure OpenAI, Anthropic, etc.) maintain their current behavior unchanged via a conservative default (
add_reasoning_to_cost: false).Test plan
add_reasoning_to_cost: truemix qualitypasses