Changelog
What's new in Voice Production AI
New voices, features, and improvements — shipped continuously. Newest releases first.
- v2.4
Studio mastering presets & faster batch
- New one-click mastering presets: Podcast, Broadcast, and Audiobook loudness targets.
- Batch generation throughput increased ~40% for Scale workspaces.
- Added MP4 export with embedded chapter markers.
- v2.3
Webhooks v2 and signed payloads
- Webhooks now sign every payload with an HMAC signature you can verify.
- Added a webhook delivery log with replay from the dashboard.
- New generation.completed and clone.ready event types.
- v2.2
40 new expressive voices
- Added 40 new preset voices, including expanded coverage for Portuguese, Hindi, and Arabic.
- Voice preview now plays inline in the voice library without spending credits.
- Improved SSML <emphasis> handling for more natural stress on long sentences.
- v2.1
Shared workspace library
- Teams can now save, tag, and reuse generated audio in a shared workspace library.
- Project-level roles: editors can generate, viewers can listen and download.
- Bulk download of a project as a zip archive.
- v2.0
Voice cloning, now studio-grade
- Rebuilt cloning pipeline: clone any voice from a ~60-second clean sample.
- Consent capture is now required at clone time and stored with the voice.
- Pro plans get up to 10 cloned voices; Scale gets up to 100.
- v1.9
AI audio enhancement
- Introduced RNNoise-based background-noise removal for uploaded and generated audio.
- Automatic loudness normalization to broadcast-ready targets.
- Added WAV export alongside MP3.