# boson-ai/higgs-audio

Text-audio foundation model from Boson AI

Repository: https://github.com/boson-ai/higgs-audio
Canonical: https://ross.abutalabs.com/products/higgs-audio
Language: Python
License: Apache-2.0
License Family: permissive
Last push: 2026-06-05T04:01:06+00:00

## Health v2 (maintenance only)
Score: 57/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 86, release rhythm 35, longevity 29
- inputs: {"age_days": 409, "days_push": 89, "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 8329, forks 640 (observed 2026-08-28T04:10:20.325291+00:00)

## What it is
Higgs Audio is a text-audio foundation model project from Boson AI providing code and weights for conversational text-to-speech with zero-shot voice cloning and prosody control. The repository now serves as the home for the legacy v2/v2.5 models, since Higgs Audio v3 is released separately as standalone weights and a hosted API.

## Use cases
- generate natural speech from text
- clone a voice from a short audio sample
- build a conversational voice assistant with expressive TTS
- self-host a TTS model on GPU with streaming output
- control emotion and prosody inline in generated speech
- synthesize speech in many languages

## When to choose
- you want open weights for expressive, conversational TTS you can self-host
- you need zero-shot voice cloning without fine-tuning
- you want multilingual speech generation with style control
- you're researching text-audio foundation models

## When to avoid
- you want the latest Higgs Audio v3 model - use the separate v3 release or hosted API instead
- you need a commercial production license without negotiating with Boson AI
- you have no GPU and don't want to use the hosted API
- you need lightweight on-device TTS

## Facets
- artifact type: library
- maturity: maintenance
- function: tts, speech-recognition, llm-inference, machine-learning
- domain: speech-processing, artificial-intelligence, large-language-models
- platform: python, self-hosted
- tags: text-to-speech, voice-cloning, audio-foundation-model, conversational-tts, multilingual, open-weights, audio, gpu, linux

## Member repositories
- boson-ai/higgs-audio (main) score 57

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:10:20.325291+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 2026-08-29T17:27:45.513252+00:00, confidence not recorded.
  - readme: https://github.com/boson-ai/higgs-audio (fetched 2026-08-28T04:10:20.325291+00:00, sha adb934fcefd3)
- Data as of 2026-08-30T08:39:29.467469+00:00.