AI TTS Microservice

Changelog

New features, improvements, and fixes.

Audio

  • Prompt field for supported models and voices - guide AI voice generation with custom instructions
  • Soft-start audio processing eliminates click artifacts
  • Safari mobile voice preview spinner fix
  • Demo voice auto-recovers from network glitches
  • Voice previews auto-recover from loading issues
  • Stop button now correctly indicates stop behavior

Text-to-Speech

  • Provider selector UI (Google enabled, Amazon Polly & Kokoro coming soon)
  • Audio format selection (WAV, MP3, OGG Opus)
  • Bitrate control for MP3 and OGG Opus output
  • History restore - recover text and settings from past generations

Sharing

  • Paid users automatically get permanent shares
  • Fixed audio format detection for shared files

Reliability

  • Fixed playback issues for history items
  • Better error recovery for failed job cleanup