jasonppy/VoiceCraft

Health v2 · maintenance only

Flags: no_releases no_license

Adoption not part of the score

8574 stars · 804 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

VoiceCraft is a token infilling neural codec language model for zero-shot speech editing and text-to-speech on in-the-wild data like audiobooks, videos, and podcasts. It can clone or edit unseen voices using only a few seconds of reference audio.

Use cases

• edit words in a podcast recording without re-recording
• clone a voice from a few seconds of reference audio
• generate speech in an unseen voice from text
• fix mispronunciations in audiobook narration
• run TTS inference from the command line or a Gradio UI
• fine-tune a TTS model on custom data

When to choose

• you need zero-shot voice cloning or speech editing on real-world recordings
• you want a research-grade TTS model you can run locally, in Colab, or via Docker
• you need to edit existing speech audio rather than synthesize from scratch

When to avoid

• you need a lightweight production TTS API without GPU requirements
• you need a permissively licensed model for commercial use (license is non-standard)
• you only need simple offline TTS without voice cloning

Facets

library · maturity active

tts speech-recognition audio-processing machine-learning deep-learning speech-processing artificial-intelligence python cross-platform zero-shot-tts speech-editing voice-cloning neural-codec-language-model gradio audio gpu docker

Member repositories

For agents

markdown · JSON · MCP: product_card(name="jasonppy/VoiceCraft")