VoiceCraft
Zero-shot speech editing and TTS using neural codec language models.
About
VoiceCraft from UT Austin is a token-infilling neural codec language model that performs both speech editing and zero-shot text-to-speech on in-the-wild audio such as audiobooks, podcasts, and videos. It can edit existing recordings or clone an unseen voice from a few seconds of reference, and ships code and demos. Inference uses a GPU with around 8 GB of VRAM. Distributed as open-source research.
Reviews (0)
Leave a Review
No reviews yet. Be the first to review!
Details
- Category
- Text-to-Speech (TTS)
- Price
- Free
- Platform
- Local/Desktop
- Difficulty
- Advanced (4/5)
- Minimum VRAM
- 8 GB
- Added
- Apr 3, 2026
Related Tools
Lightweight and expressive TTS model with 82M parameters for fast local inference.
Open-source dialogue TTS model by Nari Labs supporting multi-speaker conversations.
Pure language modeling approach to TTS without traditional audio codecs.
Transformer-based text-to-audio model by Suno that generates speech, music, and sound effects.
Multilingual TTS with zero-shot voice cloning and streaming support.
Emotion-controllable TTS engine by NetEase with 2000+ voices.