# 2noise/ChatTTS

A generative speech model for daily dialogue.

Repository: https://github.com/2noise/ChatTTS
Canonical: https://ross.abutalabs.com/products/chattts
Homepage: https://2noise.com
Language: Python
License: AGPL-3.0
License Family: copyleft
Topics: agent, text-to-speech, chat, chatgpt, chattts, chinese, chinese-language, english, english-language, gpt, llm, llm-agent, natural-language-inference, python, torch, torchaudio, tts
Last push: 2026-04-10T16:33:48+00:00

## Health v2 (maintenance only)
Score: 69/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 76, release rhythm 66, longevity 59
- inputs: {"age_days": 828, "days_push": 145, "days_rel": 145, "gap_med": 81, "n_releases_24m": 6}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 39797, forks 4256 (observed 2026-08-28T04:12:08.018571+00:00)

## What it is
ChatTTS is a generative text-to-speech model optimized for conversational dialogue scenarios such as LLM assistants, supporting English and Chinese. It offers fine-grained control over prosodic features like laughter, pauses, and interjections, with pretrained models trained on 100,000+ hours of audio.

## Use cases
- generate natural speech for chatbot dialogue
- add voice output to an LLM assistant
- synthesize Chinese and English speech with expressive prosody
- control laughter and pauses in generated speech
- build a multi-speaker conversational voice agent
- stream audio generation for real-time voice responses

## When to choose
- you need conversational, expressive TTS for a voice assistant
- your application requires Chinese or English dialogue speech with natural prosody
- you want fine-grained control over prosodic features like pauses and laughter

## When to avoid
- you need a permissively licensed model for commercial products (AGPL-3.0, academic-only release)
- you need languages other than English or Chinese
- you need production-grade TTS with speaker fine-tuning (released model lacks SFT)

## Facets
- artifact type: library
- maturity: active
- function: tts, speech-recognition, machine-learning, llm-inference
- domain: speech-processing, artificial-intelligence, large-language-models
- platform: python, cross-platform
- tags: text-to-speech, conversational-tts, chinese, english, pytorch, dialogue, voice-synthesis, prosody, natural-language-processing, gpu

## Member repositories
- 2noise/ChatTTS (main) score 69

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:12:08.018571+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 2026-08-29T16:24:06.236185+00:00, confidence not recorded.
  - readme: https://github.com/2noise/ChatTTS (fetched 2026-08-28T04:12:08.018571+00:00, sha 96d47506191a)
  - homepage: https://2noise.com (fetched 2026-08-29T07:46:38.319204+00:00, sha 6585db74bbf0)
  - registry_pypi: https://pypi.org/pypi/chattts/json (fetched 2026-08-29T07:46:38.322271+00:00, sha 4dab807b1ba1)
- Data as of 2026-08-30T08:39:29.467469+00:00.