Skip to content

feat: Add reranking instrumentation for AI SDK and Openrouter SDK#1824

Merged
Luca Forstner (lforst) merged 3 commits intomainfrom
lforst/add-rerank-tracing
Apr 15, 2026
Merged

feat: Add reranking instrumentation for AI SDK and Openrouter SDK#1824
Luca Forstner (lforst) merged 3 commits intomainfrom
lforst/add-rerank-tracing

Conversation

@lforst
Copy link
Copy Markdown
Member

Fixes #1766
Fixes #1662

@lforst Luca Forstner (lforst) marked this pull request as ready for review April 14, 2026 15:19
const normalizedModel = parseOpenRouterModelString(model);
const normalizedProvider =
(typeof provider === "string" ? provider : undefined) ||
(typeof provider === "string" ? provider.toLowerCase() : undefined) ||
Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

the provider.toLowerCase changes every openrouter span. That is fine right?

Copy link
Copy Markdown
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

yes somewhat intentional because the api returns the provider in upper case and our convention and display logic and such is lowercase

channelName: openRouterChannels.rerankRerank.channelName,
module: {
name: "@openrouter/sdk",
versionRange: ">=0.9.11 <1.0.0",
Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

opus told me that rerank does not exist in 0.9.11. Should we pick a tighter version range?

@lforst Luca Forstner (lforst) merged commit c980e07 into main Apr 15, 2026
50 checks passed
@lforst Luca Forstner (lforst) deleted the lforst/add-rerank-tracing branch April 15, 2026 08:30
@github-actions github-actions bot mentioned this pull request Apr 15, 2026
Luca Forstner (lforst) added a commit that referenced this pull request Apr 16, 2026
This PR was opened by the [Changesets
release](https://github.com/changesets/action) GitHub action. When
you're ready to do a release, you can merge this and publish to npm
yourself or [setup this action to publish
automatically](https://github.com/changesets/action#with-publishing). If
you're not ready to do a release yet, that's fine, whenever you add more
changesets to main, this PR will be updated.


# Releases
## braintrust@3.9.0

### Minor Changes

-
[#1824](#1824)
[`c980e07`](c980e07)
Thanks [@lforst](https://github.com/lforst)! - feat: Add reranking
instrumentation for AI SDK and Openrouter SDK

-
[#1738](#1738)
[`25d25d5`](25d25d5)
Thanks [@AbhiPrasad](https://github.com/AbhiPrasad)! - - feat: Add
instrumentation for @huggingface/inference
([#1807](#1807))
- feat: Add cohere-ai instrumentation
([#1781](#1781))
- fix: Capture anthropic server tool use inputs for streaming APIs
([#1776](#1776))
- feat: Capture grounding metadata for Google GenAI
([#1773](#1773))
- fix(claude-agent-sdk): Don't drop tool spans for spawning subagents
([#1779](#1779))
- feat: Track server tool use metrics for anthropic SDK
([#1772](#1772))
- fix(openai): Collect logprob and refulsals output for streaming APIs
([#1774](#1774))
- perf: Remove zod from deepCopyEvent
([#1796](#1796))
- fix(test): Double timeout for slow OpenAI API tests
([#1794](#1794))
- feat(claude-agent-sdk): Improve task lifecycle and lifecycle details
([#1777](#1777))
- ci(deps): bump actions/github-script from 8.0.0 to 9.0.0
([#1783](#1783))
- ci(deps): bump docker/setup-buildx-action from 3.12.0 to 4.0.0
([#1782](#1782))
- chore: Don't use environment (ie. github deployments) for canary tests
([#1775](#1775))
- chore: Make dependabot less annoying
([#1778](#1778))
- fix(auto-instrumentation): Upgrade @apm-js-collab/code-transformer to
v0.12.0
([#1708](#1708))
- fix(auto-instrumentation): Use sync channel for AI SDK CJS
streamText/streamObject in v4+
([#1768](#1768))
- fix: Give AI SDK top-level api spans type function
([#1769](#1769))

-
[#1814](#1814)
[`d9b9923`](d9b9923)
Thanks [@lforst](https://github.com/lforst)! - - feat: Add per-input
trialCount support to Eval()

-
[#1821](#1821)
[`1b5de11`](1b5de11)
Thanks [@lforst](https://github.com/lforst)! - feat: Instrument Google
GenAI embedContent for text

### Patch Changes

-
[#1825](#1825)
[`f389036`](f389036)
Thanks [@lforst](https://github.com/lforst)! - fix(ai-sdk): Restore
prompt cache metrics

-
[#1813](#1813)
[`2434a0e`](2434a0e)
Thanks [@lforst](https://github.com/lforst)! - fix(openai-agents): End
child spans on trace end

-
[#1836](#1836)
[`5581357`](5581357)
Thanks [@stretpjc](https://github.com/stretpjc)! - feat: Add
x-bt-use-gateway header to allowed CORS headers

---------

Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
Co-authored-by: Luca Forstner <luca.forstner@gmail.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

[BOT ISSUE] OpenRouter rerank.rerank() not instrumented Vercel AI SDK rerank() not instrumented (wrapper and auto-instrumentation)

2 participants