phonon-wasm

phonon-wasm

Turns text into speech with the Phonon 31M model running on your own CPU inside the browser, streaming audio as it is computed. For people who want speech from text without a server.

Made by fusorlabs 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: phonon-wasm is a text-to-speech page that runs the Phonon 31M model inside your browser. The model is compiled to WebAssembly with fusor-ai and works on your own CPU. It downloads once, then runs sealed from the network and streams each 80 ms frame to your speakers as it is computed.

Who it suits: People who want text read aloud without a server doing the work. It also suits developers who want to see a small speech model compiled to WebAssembly and served as a static page.

Industry | Who uses it | How they use it | Example
Content creation | Writers and video makers | Hear a draft read aloud | Type a short passage and play it back
Accessibility | People who prefer listening | Have text read to them on their own machine | Paste a paragraph and listen
Software | Developers | Study a model compiled to WebAssembly | Read the engine and model folders in a local copy

If you need a voice other than the one this model gives, you can skip it.

Where to use it: The page works in these places:

• A web browser: open the Space page on Hugging Face and use it there.
• Windows, macOS or Linux: download the files and open index.html in a browser.

Use case | Who | What they get back
Hear text read aloud | Anyone | Speech made on your own CPU
Use it without a server | Privacy-minded users | A model that runs with no network use once downloaded
Study the build | Developers | A static page with engine and model folders

How to set it up: Using the page online needs nothing. For a local copy you need a few things:

• Python with pip, to fetch the files
• The huggingface_hub package, which the first command installs
• A web browser

The steps download the Space, which includes its model folder, and then open the main page by hand.

1. Install the Hugging Face package.
2. Download the Space into a folder called phonon-wasm.
3. Move into that folder.
4. Open index.html in your browser.

pip install -U huggingface_hub
hf download fusorlabs/phonon-wasm --repo-type space --local-dir phonon-wasm
cd phonon-wasm

How to use it: Open the page, let the model finish its one-time download, then give it text and listen to the speech it plays. The audio is streamed as it is computed.

Tool | What it does
index.html | The page that loads the model and plays the speech
hf download | Copies the Space's files into a folder on your computer

Example: You open the page and wait for the model to download, which happens only once. You then type a sentence, and the page starts speaking it while the rest is still being computed on your CPU.

For newcomers and experts: If you are new to this, open the page, give it a short sentence and listen. You need no technical background. The first visit takes longer because of the one-time model download.

If you know the field, the Space names the model gradium/phonon-31m-en, which by its name is an English model. It is compiled to WebAssembly with fusor-ai and runs on the CPU, with no GPU and no backend. Audio goes out in 80 ms frames as each one is computed.

You open the static page
The model downloads once
The model runs on your CPU in WebAssembly
Each 80 ms frame streams to your speakers

Good to know: The card does not give the model's download size or the languages it covers.

• License: mit, as the Space lists it.
• Cost to run: none named.
• Data it reaches: the one-time model download. After that, the model runs sealed from the network.
• Not for: none stated in the card.

On Hugging Face, 13 November 2026

• Likes: 2
• Built with: static
• License: mit

Links

• Space: https://huggingface.co/spaces/fusorlabs/phonon-wasm
• Files: https://huggingface.co/spaces/fusorlabs/phonon-wasm/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.