Breeze TTS 2 needs about 4.8 GB of memory at Q8_0 (a 3.9 GB file) with an 8K context window, so it fits on a 6 GB GPU. This repository has 5 quants, from 2.2 GB (Q4_0) to 7.0 GB (BF16).
Breeze TTS 2 – real‑time bilingual text‑to‑speech
Breeze TTS 2 is a 3.5B bilingual (English/Chinese) model for real‑time TTS. It supports plain TTS, voice design, and voice clone with ultra‑low latency and streaming. The model is open‑weight and requires a CUDA‑capable NVIDIA GPU.
Good at
- Real‑time streaming
- Ultra‑low latency TTFA
- Voice design and clone
Limits
- Requires 12 GB GPU for eager inference
- Not compatible with llama.cpp
- No vision support
Best for
- Live streaming audio generation
- Interactive voice assistants
- Voice cloning for media
Who it’s for. Developers building real‑time TTS applications or researchers needing a bilingual, open‑weight model. The model is suitable for non‑commercial use under its license.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q4_0 | 2.2 GB | 2.9 GB | Fits | Get file → | |
| — | 3.8 GB | 4.8 GB | Fits | Get file → | |
| Q8_0 | 3.9 GB | 4.8 GB | Fits | Get file → | |
| F16 | 7.0 GB | 8.3 GB | Too big | Get file → | |
| BF16 | 7.0 GB | 8.3 GB | Too big | Get file → |
For a 12 GB GPU, the Q4_0 variant (2.18 GB file, 2.9 GB memory) is a good fit. If more memory is available, Q8_0 (3.85 GB file, 4.8 GB memory) offers higher precision. Larger variants (F16, BF16) require about 8.3 GB memory and are suitable only on GPUs with 24 GB or more.
What GPU do you need for Breeze TTS 2?
These figures add the file size, an fp16 KV cache at an 8K context window and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.
- 4 GB VRAM: the best quant that fits is Q4_0 (2.2 GB file, about 2.9 GB in use).
- 6 GB VRAM: the best quant that fits is Q8_0 (3.9 GB file, about 4.8 GB in use).
- 8 GB VRAM: the best quant that fits is Q8_0 (3.9 GB file, about 4.8 GB in use).
- 12 GB VRAM: the best quant that fits is BF16 (7.0 GB file, about 8.3 GB in use).
Which Breeze TTS 2 quant should you choose?
Q8_0 (3.9 GB) is the middle option in this repository. Close to lossless. It costs roughly twice the memory of a 4-bit file, so it suits cases where you have plenty of VRAM.
If memory is tight, the smallest file is Q4_0 at 2.2 GB, with the quality trade-offs that come with fewer bits. The usual sweet spot between file size and quality, and the quant most people start with.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about Breeze TTS 2
How much VRAM does Breeze TTS 2 need?
About 4.8 GB at Q8_0 with an 8K context window. That is the 3.9 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.
Can I run Breeze TTS 2 on an 8 GB GPU?
Yes, with Q8_0: the 3.9 GB file needs about 4.8 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of Breeze TTS 2 GGUF?
It depends on the quant: Q4_0 is 2.2 GB, file is 3.8 GB, Q8_0 is 3.9 GB, F16 is 7.0 GB, BF16 is 7.0 GB.
Where can I download Breeze TTS 2 GGUF?
The files are on Hugging Face in the EvoAwaken-Workshop/Breeze-TTS-2-gguf repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does Breeze TTS 2 use?
Hugging Face lists the license as other. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.
Does Breeze TTS 2 support voice design?
Yes. Use the --instruction flag to describe voice characteristics and set --cfg-scale for control.
Is the model compatible with llama.cpp?
No. Breeze and breeze_audio architectures are specific to the Rust implementation and not supported by llama.cpp.