Vevo for Zero-shot VC, TTS, and More

Vevo for Zero-shot VC, TTS, and More

Imitates a voice from a reference clip without training on that speaker, for voice conversion and text-to-speech. A demo of the Vevo model from the Amphion project, for researchers and audio developers.

Made by amphion on Hugging Face. These notes are Mindket's, written from its Space page and README; the app's own files and license are on Hugging Face.

About

What it is: This Space demos Vevo, a model for controllable zero-shot voice imitation. Zero-shot means it can copy a voice it was never trained on, using only a reference sample. The title lists voice conversion and text-to-speech among its uses.

Who it suits: It suits speech researchers and audio developers who want to hear what a current voice-imitation model produces. It also suits students who are learning how voice conversion and text-to-speech work.

Industry | Who uses it | How they use it | Example
Research | Speech researchers | Listen to a published voice-imitation model | Compare the output with a reference clip
Software | Audio developers | Judge whether the model fits a product idea | Try voice conversion on a short sample
Education | Students | Learn what zero-shot voice cloning means | Run text-to-speech and listen to the result
Media | Audio producers | Test voice imitation on their own recordings | Convert a clip they recorded themselves

Anyone who needs a ready, production voice service can skip it, since this is a research demo.

Where to use it: The demo runs in two places.

• A web browser, on the page of the Space at Hugging Face
• Windows, macOS or Linux, as a local copy run with Python

Use case | Who | What they get back
Voice conversion | Researchers | Speech that carries the style of a reference voice
Text-to-speech | Developers | Spoken audio made from typed text
Model demo | Students | A hands-on look at zero-shot voice imitation

How to set it up: The online page needs nothing installed. A local copy needs the following.

• Python and pip
• Windows, macOS or Linux
• The packages listed in requirements.txt

1. Install Python.
2. Install the Hugging Face hub package and download a copy of the Space.
3. Move into the new folder and install Gradio 5.25.2 and the packages in requirements.txt.
4. Start the app with python app.py.

pip install -U huggingface_hub
hf download amphion/Vevo --repo-type space --local-dir Vevo
cd Vevo
pip install gradio==5.25.2
pip install -r requirements.txt
python app.py

How to use it: Open the page in a browser, give it a reference voice and the content to speak, and listen to the result. On a local copy, this command starts the page.

python app.py

Tool | What it does
Vevo | The model the Space names, which imitates a voice in a zero-shot way
app.py | Holds the Gradio page
requirements.txt | Lists the Python packages the app needs

Example: You supply a short recording of a voice and a sentence of text. The page answers with audio of that sentence spoken in a voice that imitates your reference.

For newcomers and experts: If you are new to this, start with a clear, short reference recording and a short sentence. Listen to the result, then change the reference and hear how the voice changes.

If you know the field, the Space is a front end for the amphion/Vevo model on a ZeroGPU Space with A10G hardware. The page gives no architecture notes, so the model's own documentation is the place to look for training data, limits and the meaning of its controls.

Reference voice sample
Text or source speech
Vevo imitates the voice
Audio comes back

Good to know: The Space page holds only a one-line description. It lists no usage rules, so use only voices you have permission to copy.

• License: MIT
• Cost to run: none named
• Data it reaches: the audio and text you give it, and the amphion/Vevo model
• Hardware: ZeroGPU with an A10G at Hugging Face

On Hugging Face, 13 November 2026

• Likes: 104
• Built with: gradio
• License: mit

Links

• Space: https://huggingface.co/spaces/amphion/Vevo
• Files: https://huggingface.co/spaces/amphion/Vevo/tree/main

Try it

Open the Space in your browser, or copy it to your own Hugging Face account from its page. Nothing in this kit runs by itself.