# metavoiceio/metavoice-src

Foundational model for human-like, expressive TTS

Repository: https://github.com/metavoiceio/metavoice-src
Canonical: https://ross.abutalabs.com/products/metavoice-src
Homepage: https://themetavoice.xyz/
Language: Python
License: Apache-2.0
License Family: permissive
Topics: text-to-speech, ai, deep-learning, pytorch, speech, speech-synthesis, tts, voice-clone, zero-shot-tts
Last push: 2024-07-30T22:13:01+00:00

## Health v2 (maintenance only)
Score: 26/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 67
- inputs: {"age_days": 939, "days_push": 764, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 4205, forks 692 (observed 2026-08-28T04:08:39.021901+00:00)

## What it is
MetaVoice-1B is a 1.2B parameter foundational text-to-speech model trained on 100K hours of speech, focused on emotional rhythm and tone in English. It supports zero-shot voice cloning with 30 seconds of reference audio and cross-lingual cloning via finetuning, released under Apache 2.0.

## Use cases
- generate human-like expressive speech from text
- clone a voice from 30 seconds of reference audio
- synthesize arbitrary-length English text with emotional tone
- finetune TTS for cross-lingual voice cloning with 1 minute of data
- self-host a TTS server with an API and web UI

## When to choose
- you need expressive, emotionally natural English TTS
- you want zero-shot voice cloning under a permissive Apache 2.0 license
- you have a GPU with at least 12GB VRAM and want to self-host

## When to avoid
- you need production TTS for many languages without finetuning
- you have no GPU available
- you need actively maintained software with recent updates

## Facets
- artifact type: library
- maturity: maintenance
- function: tts, machine-learning, deep-learning, speech-recognition
- domain: speech-processing, artificial-intelligence, deep-learning
- platform: python, self-hosted
- tags: text-to-speech, voice-cloning, zero-shot-tts, pytorch, speech-synthesis, expressive-tts, audio, gpu, linux, docker

## Member repositories
- metavoiceio/metavoice-src (main) score 26

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:39.021901+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T18:22:31.442433+00:00, confidence not recorded.
  - readme: https://github.com/metavoiceio/metavoice-src (fetched 2026-08-28T04:08:39.021901+00:00, sha 97b9edd9e6b4)
- Data as of 2026-08-30T08:39:29.467469+00:00.