VibeVoice
github.com/microsoft/vibevoice- Category
- AI Tools
- Pricing
- Open Source
- Type
- TOOL
- Builder
- @microsoft
- GitHub
- 50.5k stars
- Added
- Apr 12, 2026
About
Open-source voice AI framework that includes advanced speech recognition (ASR) for 60-minute audio transcription with speaker diarization, text-to-speech (TTS) for 90-minute multi-speaker synthesis, and real-time streaming TTS. Operates at ultra-low 7.5Hz frame rate for efficient long-form audio processing.
Why it made the leaderboard
Microsoft Research's open-source voice family built on 7.5Hz continuous speech tokenizers: ASR that transcribes 60 minutes in a single pass with speaker diarization, TTS that sustains 90-minute multi-speaker synthesis, and real-time streaming TTS. Long-form audio without the chunking hacks.
Tags
voice-aispeech-recognitiontext-to-speechopen-sourcelong-form-audiospeaker-diarizationstreaming-ttsmicrosoft
Tech Stack
Python
Featured in Intel
Comments (0)
No comments yet
Indexed by a proprietary survey. Corrections welcome.