fix(mastra): record time_to_first_token for streaming model spans#2214
Open
Cedric / ViaDézo1er (viadezo1er) wants to merge 1 commit into
Open
fix(mastra): record time_to_first_token for streaming model spans#2214Cedric / ViaDézo1er (viadezo1er) wants to merge 1 commit into
Cedric / ViaDézo1er (viadezo1er) wants to merge 1 commit into
Conversation
The Mastra observability exporter now derives time_to_first_token (in seconds) from a model span's completionStartTime attribute and its start time — the metric Braintrust surfaces for streaming LLM calls — matching how our other wrappers emit TTFT. The raw completionStartTime is no longer duplicated into span metadata. Token usage mapping (prompt/completion/total, cached, cache-creation, and reasoning tokens) was already at parity with @mastra/braintrust and is unchanged. Ports the TTFT half of mastra-ai/mastra#11029 (fixes #10174, #9821, #9853). Adds js/src/wrappers/mastra.test.ts and regenerates the mastra-instrumentation e2e snapshots (streaming model spans now carry time_to_first_token).
Cedric / ViaDézo1er (viadezo1er)
force-pushed
the
cedric/token-usage-ttft
branch
from
8f9432f to
656a2e1
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
The Mastra observability exporter now derives time_to_first_token (in
seconds) from a model span's completionStartTime attribute and its start
time — the metric Braintrust surfaces for streaming LLM calls — matching
how our other wrappers emit TTFT. The raw completionStartTime is no
longer duplicated into span metadata.
Token usage mapping (prompt/completion/total, cached, cache-creation,
and reasoning tokens) was already at parity with @mastra/braintrust and
is unchanged.
Ports the TTFT half of mastra-ai/mastra#11029 (fixes #10174, #9821,
#9853). Adds js/src/wrappers/mastra.test.ts and regenerates the
mastra-instrumentation e2e snapshots (streaming model spans now carry
time_to_first_token).