Pocket TTS (CPU) by Xlnk

Pocket TTS (CPU) by Xlnk

A Gradio page for PocketTTS, a light text-to-speech model from Kyutai Labs that runs on a CPU, with several built-in voices. For anyone who wants to turn text into speech without a GPU.

Made by Xlnk 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 is a Gradio web interface for PocketTTS, a lightweight text-to-speech model from Kyutai Labs with about 100 million parameters. It is tuned to run on a CPU. The interface was built by a separate author and is not officially tied to Kyutai Labs.

Who it suits: People who want to hear text read aloud without a GPU, and developers who want a small example of a Gradio front end for a speech model.

Industry | Who uses it | How they use it | Example
Content creation | Video makers | Get a quick voice track for a draft | Type a script and pick a voice
Software development | Developers | Study a small Gradio speech front end | Read app.py in the repository
Accessibility | App builders | Prototype read-aloud features on a CPU | Try a sentence in each voice

People who need to clone a specific voice can skip it, because it offers a fixed set of built-in voices.

Where to use it:

• A browser: the Space runs on Hugging Face with nothing to install
• Windows, macOS or Linux: run a copy of the Space with Python

Use case | Who | What they get back
Read text aloud | Anyone | Speech in one of the built-in voices
Draft a voice track | Video makers | A quick spoken version of a script

How to set it up:

• Python with pip on Windows, macOS or Linux, 3.10 or newer
• The Python packages listed in requirements.txt, which the README says include PyTorch 2.5 or newer
• Nothing at all if you only use the Space in a browser

1. Install the Hugging Face hub package.
2. Download the Space's files into a folder.
3. Move into that folder.
4. Install Gradio 6.3.0 and the packages from requirements.txt.
5. Start the app.

pip install -U huggingface_hub
hf download Xlnk/tts --repo-type space --local-dir tts
cd tts
pip install gradio==6.3.0
pip install -r requirements.txt
python app.py

How to use it: Open the Space page in a browser, or start your own copy with the command below. Type your text, choose a voice and run it.

python app.py

The built-in voices are these.

• alba: the default voice
• marius, javert and jean
• fantine, cosette, eponine and azelma

Example: You type a short sentence and keep the default voice, alba. The app returns the sentence as speech, which you can listen to in the browser.

For newcomers and experts: If you are new, open the Space and try a short sentence in the default voice, then switch voices and listen again. Nothing needs installing.

For experts, the README says the model is tuned for CPUs and that the first audio chunk arrives in about 200 milliseconds on local installs. The Space itself runs on cpu-basic hardware, so a hosted run may differ.

You type some text
You choose a built-in voice
PocketTTS runs on the CPU
You hear the speech

Good to know:

• License: apache-2
• Cost to run: none named
• Data it reaches: the text you type
• Not for: official Kyutai Labs use, since the interface is an independent demo

On Hugging Face, 13 November 2026

• Likes: 7
• Built with: gradio
• License: apache-2.0

Links

• Space: https://huggingface.co/spaces/Xlnk/tts
• Files: https://huggingface.co/spaces/Xlnk/tts/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.