fix(langchain): normalize anthropic cache token metrics#411
Merged
Abhijeet Prasad (AbhiPrasad) merged 1 commit intoMay 11, 2026
Merged
Conversation
LangChain Anthropic usage reports cache read/write tokens separately from raw input and total counts, which could produce Braintrust spans where cache creation tokens exceeded total tokens. Fold cache read and creation tokens into prompt_tokens and total_tokens when LangChain appears to expose uncached prompt totals. This keeps Braintrust token metrics internally consistent for users relying on prompt caching: before, spans could underreport prompt/total tokens; after, prompt_tokens includes uncached, cached, and cache-creation tokens and total_tokens remains prompt plus completion tokens. Extend the existing VCR prompt-caching regression test to assert the normalized metric convention for cache creation and cache reads.
Luca Forstner (lforst)
approved these changes
May 11, 2026
Abhijeet Prasad (AbhiPrasad)
deleted the
abhi-fix-bt-5150-langchain-cache-tokens
branch
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
LangChain Anthropic usage reports cache read/write tokens separately from raw input and total counts, which could produce Braintrust spans where cache creation tokens exceeded total tokens.
Fold cache read and creation tokens into prompt_tokens and total_tokens when LangChain appears to expose uncached prompt totals. This keeps Braintrust token metrics internally consistent for users relying on prompt caching: before, spans could underreport prompt/total tokens; after, prompt_tokens includes uncached, cached, and cache-creation tokens and total_tokens remains prompt plus completion tokens.
Extend the existing VCR prompt-caching regression test to assert the normalized metric convention for cache creation and cache reads.
fixes https://linear.app/braintrustdata/issue/BT-5150/langchain-callback-does-not-normalize-anthropic-cache-tokens-into