feat(whispering): send no_verbatim for ElevenLabs scribe_v2#1725
Draft
osman-samil wants to merge 1 commit into
Draft
feat(whispering): send no_verbatim for ElevenLabs scribe_v2#1725osman-samil wants to merge 1 commit into
osman-samil wants to merge 1 commit into
Conversation
The bundled `elevenlabs@1.59.0` SDK predates the scribe_v2 release and does not expose the `no_verbatim` flag in its `BodySpeechToTextV1SpeechToTextPost` type. Upgrading to `@elevenlabs/elevenlabs-js@2.x` is a package rename plus breaking-change migration, so this patch keeps the dependency footprint unchanged and posts the multipart form to `/v1/speech-to-text` directly. When the user picks scribe_v2, we now send `no_verbatim=true`, which strips filler words, false starts, and non-speech sounds at the model. This is a no-op on scribe_v1 / scribe_v1_experimental (parameter is silently ignored), so existing flows are unaffected. Ref: https://elevenlabs.io/docs/api-reference/speech-to-text/convert Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
What
When the user selects ElevenLabs
scribe_v2, sendno_verbatim=trueto/v1/speech-to-text. ElevenLabs strips filler words, false starts, and non-speech sounds at the model, so the output is dictation-ready without a downstream transformation step.No-op on
scribe_v1/scribe_v1_experimental(the parameter is ignored when not supported), so existing flows are unaffected.Ref: https://elevenlabs.io/docs/api-reference/speech-to-text/convert
Why direct fetch
The bundled SDK is the legacy
elevenlabs@1.59.0package, which predatesscribe_v2and exposes neither the new model id norno_verbatimin itsBodySpeechToTextV1SpeechToTextPosttype. Upgrading to@elevenlabs/elevenlabs-js@2.xis a package rename plus a breaking-change migration. To keep the diff small and avoid touching the lockfile, this patch posts the multipart form directly. The other adapters in this folder follow the same multipart pattern (seegroq.ts,openai.ts), so this is consistent with the rest of the cloud transcription services.Notes
ApiErrorvariant so non-2xx responses surface the status and body instead of being flattened intoUnexpected.Draft for now while I confirm the build artifact runs cleanly on Windows; happy to mark ready or split into a settings-toggle follow-up if maintainers prefer opt-in.
Need help on this PR? Tag
@codesmithwith what you need.