---
license: apache-2.0
base_model: FireRedTeam/FireRedTTS3
pipeline_tag: text-to-speech
library_name: comfyui
tags:
- tts
- text-to-speech
- voice-cloning
- voice-design
- speech-editing
- fireredtts
- qwen3
- comfyui-custom-node
- bf16
language:
- zh
- en
- yue
- ja
- ko
- es
- fr
- ru
- ar
- tr
- id
- pt
- it
- nl
- vi
- de
- uk
- th
- pl
- ro
- el
- cs
- fi
- hi
---

# FireRedTTS3-bf16 (community mixed-precision mirror)

Half-size copy of [FireRedTeam/FireRedTTS3](https://huggingface.co/FireRedTeam/FireRedTTS3) for
[FireRedTTS3-ComfyUI](https://github.com/Saganaki22/FireRedTTS3-ComfyUI).

Conversion policy - only the tensors the official inference code already computes under bf16 autocast
are stored in bf16; everything that officially runs in fp32 stays fp32:

| Component | bf16 | fp32 |
| --- | --- | --- |
| `fireredtts3_base` | Qwen3 backbone LLM | PatchEncoder, DiT flow head, stop head, speaker projections |
| `fireredtts3_instruct` | Qwen3 backbone LLM | PatchEncoder, DiT flow head, stop head |
| `redae` | audio encoder | decoder + ISTFT head |
| `campp` | - | everything |

Same-seed generation against the official fp32 weights produces identical waveforms
(cosine 1.0000, SNR > 80 dB) because the stored bf16 values round exactly like the
runtime autocast cast does. `campp/`, `text_tokenizer/`, and `fasttext/lid.176.ftz` are
copied through unchanged; only the `dtype` field in each `config.json` changed.

## Usage Disclaimer

- The project incorporates zero-shot voice cloning functionality; Please note that this capability is intended **solely for academic research purposes**.
- **DO NOT** use this model for **ANY illegal activities**❗️❗️
- The developers assume no liability for any misuse of this model.
- If you identify any instances of **abuse**, **misuse**, or **fraudulent** activities related to this project, **please report them to our team immediately.**


## Citation

```bib
@article{fireredtts3,
  title   = {FireRedTTS3: Unified Speech Generation and Editing with Semantically Enriched Speech Representations},
  author  = {FireRed Team},
  journal = {arXiv preprint},
  year    = {2026},
}
```


## Acknowledgements

- [Qwen3](https://github.com/QwenLM/Qwen3) and [Qwen2-Audio](https://github.com/QwenLM/Qwen2-Audio) for the language model and audio understanding foundations
- [DiTAR](https://arxiv.org/abs/2502.03930) for the patch-level diffusion autoregressive formulation
- [X-Codec](https://github.com/zhenye234/xcodec) for the discriminator design used in RedAE training
- [CAM++](https://modelscope.cn/models/iic/speech_campplus_sv_en_voxceleb_16k) for speaker embedding extraction
- [fastText](https://fasttext.cc/docs/en/language-identification.html) for automatic language identification
- [WeTextProcessing](https://github.com/wenet-e2e/WeTextProcessing) (wetext) for the Chinese / English text normalization front-end


All credit to the FireRed Team - see the upstream repo and model card. Apache-2.0.