[GitHub Trending] microsoft/VibeVoice
7.3 relevance
Score Breakdown
technical depth 7
novelty 8
actionability 6
community 7
strategic 8
personal 9
Scored daily by a customisable AI persona to surface the most relevant engineering leadership news.
Microsoft's open-source voice AI, high strategic impact and strong alignment with AI/ML interests.
Summary
Microsoft open-sourced VibeVoice, a family of frontier voice AI models including VibeVoice-ASR for long-form speech recognition and VibeVoice-Realtime for streaming TTS. The ASR model handles 60-minute audio in a single pass with speaker diarization and timestamps, compressed via heterogeneous quantization (I8_S + I2_S) from 4.62 GB to 1.58 GB for real-time CPU inference. VibeVoice uses continuous speech tokenizers at 7.5 Hz and a next-token diffusion framework with an LLM for context understanding.
Author
microsoft