# FireRedTeam/FireRedTTS2

Long-form streaming TTS system for multi-speaker dialogue generation

Repository: https://github.com/FireRedTeam/FireRedTTS2
Canonical: https://ross.abutalabs.com/products/fireredtts2
Language: Python
License: Apache-2.0
License Family: permissive
Last push: 2025-10-26T11:30:01+00:00

## Health v2 (maintenance only)
Score: 40/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 49, release rhythm 35, longevity 26
- inputs: {"age_days": 365, "days_push": 311, "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 1428, forks 128 (observed 2026-08-28T04:04:42.052556+00:00)

## What it is
FireRedTTS-2 is a long-form streaming text-to-speech system for multi-speaker dialogue generation, built in PyTorch with a dual-transformer architecture on a 12.5Hz streaming speech tokenizer. It supports zero-shot voice cloning, multilingual synthesis, and low first-packet latency (~140ms on an L20 GPU).

## Use cases
- generate podcast audio with multiple speakers from a script
- build a chatbot with natural streaming voice responses
- clone a voice with zero-shot samples for cross-lingual speech
- create synthetic ASR training data with random timbres
- generate long multi-speaker dialogues in Chinese, English, Japanese, Korean, French, German, or Russian

## When to choose
- you need low-latency streaming TTS for conversational agents or chatbots
- you want multi-speaker dialogue generation with stable speaker switching
- you need zero-shot voice cloning across multiple languages
- you want to synthesize long-form podcast-style audio

## When to avoid
- you need a lightweight CPU-only TTS for embedded devices
- you only need simple single-phrase TTS without conversational context
- you require non-PyTorch deployment runtimes like ONNX or TensorRT out of the box

## Facets
- artifact type: library
- maturity: active
- function: tts, speech-recognition, machine-learning, llm-inference
- domain: speech-processing, artificial-intelligence
- platform: python
- tags: voice-cloning, podcast-generation, streaming-tts, multi-speaker-dialogue, zero-shot-voice-cloning, pytorch, audio, natural-language-processing, gpu, linux

## Member repositories
- FireRedTeam/FireRedTTS2 (main) score 40

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:42.052556+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 2026-08-30T04:37:17.598359+00:00, confidence not recorded.
  - readme: https://github.com/FireRedTeam/FireRedTTS2 (fetched 2026-08-28T04:04:42.052556+00:00, sha 971d9c669211)
- Data as of 2026-08-30T08:39:29.467469+00:00.