zipvoice needs about 0.7 GB of memory at Q8_0 (a 0.2 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 2 quants, from 0.2 GB (Q8_0) to 0.6 GB.
ZipVoice: Zero‑shot text‑to‑speech with flow matching
ZipVoice is a non‑autoregressive TTS model that generates high‑quality speech from text. It uses a flow‑matching backbone and a Vocos vocoder to produce 24 kHz mono WAV output. The model is packaged for audio.cpp and released under Apache‑2.0.
Good at
- Fast, zero‑shot voice cloning
- High‑quality 24 kHz mono output
- Supports English and Chinese text
Limits
- No vision support
- Requires 8 k context for memory estimates
- Only weight‑only quantization available
Best for
- Voice‑cloning demos
- Rapid TTS prototyping
- Embedded TTS in audio.cpp projects
Who it’s for. Developers building TTS applications who need a lightweight, zero‑shot model. The model is suitable for projects that can run audio.cpp and require high‑quality speech synthesis without fine‑tuning.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q8_0 | 0.2 GB | 0.7 GB | Fits | Get file → | |
| — | 0.6 GB | 1.1 GB | Fits | Get file → |
For 8 GB VRAM, the Q8_0 file (0.16 GB) is recommended, using about 0.7 GB of VRAM at 8 k context. If memory is limited, the unknown quant file (0.55 GB) uses about 1.1 GB of VRAM. Both are weight‑only quantizations.
What GPU do you need for zipvoice?
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 the main file (0.6 GB file, about 1.1 GB in use).
Which zipvoice quant should you choose?
Q8_0 (0.2 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.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about zipvoice
How much VRAM does zipvoice need?
About 0.7 GB at Q8_0 with an 8K context window. That is the 0.2 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 zipvoice on an 8 GB GPU?
Yes, with the smaller files: the 0.6 GB file needs about 1.1 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of zipvoice GGUF?
It depends on the quant: Q8_0 is 0.2 GB, file is 0.6 GB.
Where can I download zipvoice GGUF?
The files are on Hugging Face in the davidxifeng/zipvoice-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 zipvoice use?
Hugging Face lists the license as apache-2.0. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.
Can I run ZipVoice on 8 GB VRAM?
Yes. The Q8_0 quantization uses about 0.7 GB of VRAM at 8 k context, well within an 8 GB budget.
Does ZipVoice support multi‑speaker TTS?
No. ZipVoice is a single‑speaker model; ZipVoice‑Dialog is the multi‑speaker variant.
What license applies to the model weights?
The weights are released under Apache‑2.0, with a CC‑BY‑NC restriction on the training data.