Skip to content

feat(inference): STT diarization capabilities and speaker_id on TimedString, add xAI TTS support for inference#5438

Merged
russellmartin-livekit merged 8 commits intomainfrom
diarization-from-main
Apr 16, 2026
Merged

feat(inference): STT diarization capabilities and speaker_id on TimedString, add xAI TTS support for inference#5438
russellmartin-livekit merged 8 commits intomainfrom
diarization-from-main

Conversation

@russellmartin-livekit
Copy link
Copy Markdown
Contributor

Restore inference STT speaker_id passthrough, TimedString.speaker_id, and diarization capability detection from extra_kwargs (diarize / speaker_labels), with tests.

The other PR was corrupted by my git fiddling
#5283

…String

Restore inference STT speaker_id passthrough, TimedString.speaker_id, and
diarization capability detection from extra_kwargs (diarize / speaker_labels),
with tests.
@russellmartin-livekit russellmartin-livekit requested review from a team April 13, 2026 23:24
@russellmartin-livekit russellmartin-livekit self-assigned this Apr 13, 2026
Copy link
Copy Markdown
Contributor

@devin-ai-integration devin-ai-integration Bot left a comment

Choose a reason for hiding this comment

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

✅ Devin Review: No Issues Found

Devin Review analyzed this PR and found no potential bugs to report.

View in Devin Review to see 4 additional findings.

Open in Devin Review

…g. Added detailed attributes to TimedString class and refactored diarization capability checks in STT to streamline provider-specific handling. Updated tests to verify new functionality.
…' attribute for Inverse Text Normalization in STT.
…ins-runway to the project. Update dependencies and extras in uv.lock to include new plugins and their requirements.
@russellmartin-livekit russellmartin-livekit changed the title feat(inference): STT diarization capabilities and speaker_id on TimedString feat(inference): STT diarization capabilities and speaker_id on TimedString, add xAI TTS support for inference Apr 16, 2026
@russellmartin-livekit russellmartin-livekit merged commit 31b4756 into main Apr 16, 2026
24 checks passed
@russellmartin-livekit russellmartin-livekit deleted the diarization-from-main branch April 16, 2026 20:47
Copy link
Copy Markdown
Contributor

👋 This is an automated Claude Code Routine created by @toubatbrian. Right now it is in experimentation stage. The automation will start porting this PR into agents-js automatically.

What will be ported:

  • speaker_id field on TimedString and SpeechData types
  • diarization capability flag on STTCapabilities
  • Diarization capability detection from provider-specific options (diarize, speaker_labels)
  • xAI STT model type support in the inference module
  • speaker_id passthrough in speech data building

What will NOT be ported (Python-specific):

  • uv.lock changes (Python lockfile)
  • Python-specific type overloads and TypedDict definitions (will be adapted to TypeScript equivalents)

Generated by Claude Code

toubatbrian pushed a commit to livekit/agents-js that referenced this pull request Apr 16, 2026
Port of livekit/agents#5438 — adds STT diarization capability detection
and speaker_id passthrough from the Python agents framework.

https://claude.ai/code/session_01VtE2b4qcjcN21cvDhsdcFo
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.

3 participants