Phonon voices

Phonon voices

Provides 102 ready-made voices for the Phonon S text-to-speech model across English, French, German, Spanish and Portuguese, each as a small embedding file you load directly, with no cloning.

Made by gradium on Hugging Face. These notes are Mindket's, written from its model page and model card; the model's own files and license are on Hugging Face.

About

What it is: Phonon voices is a set of 102 voices for Phonon S, a 31M text-to-speech model, made by gradium. It holds 90 catalogue voices in 7 locales and 12 English character voices. Each voice ships as one safetensors embedding file that Phonon loads as it is, so you choose from a list and nothing is cloned.

Who it suits: Developers who build speech features with Phonon S and want a choice of voices in several languages.

Industry | Who uses it | How they use it | Example
Voice apps | App developers | Give an assistant a chosen voice | A Spanish assistant with a male voice
Media | Video and audio creators | Narrate scripts in a set voice | A French narration for a short video
Education | Language teachers | Read lessons in several languages | A German or Portuguese reading
Games | Game developers | Use character voices in English | A character voice for a game figure

If you do not use Phonon S, you can skip it, since the files are made for that model.

Where to use it:

• Phonon S, the 31M text-to-speech model these voices were made for
• Python on Windows, macOS or Linux, to download the files
• The card's own page, where each voice has a sample you can play

Use case | Who | What they get back
Narration | Creators | Speech in the voice you chose, in one of five languages
Assistants | App developers | A fixed voice for every reply
Character speech | Game developers | One of 12 English character voices

How to set it up:

• Python and pip, to fetch the files
• Phonon S, which loads the voice files
• Disk space for the voice files you want

1. Install the Hugging Face hub tools.
2. Download the voice files.
3. Point Phonon S at the voice file you choose, following Phonon's own instructions.

pip install -U huggingface_hub
hf download gradium/phonon-voices

How to use it: Browse the voice tables on the card, which list each voice by name and ID, and play the sample for each one. Pick a voice and load its file in Phonon. The card shows no run command, so follow the Phonon documentation for loading a voice.

Example: You listen to the sample of a voice called Zoey in the US English list, decide you like it, and load her file in Phonon S. The text you give Phonon is then spoken in that voice.

For newcomers and experts: If you are new to this, play a few samples on the card page and pick one before you download anything. The files are small embeddings, not full models.

Each file is a safetensors embedding in the format ba7a5a45@300, with one float32 tensor named speaker_wavs of shape 1 by 512 by 125. Phonon uses it directly with no cloning step. The set covers 48 female voices, 42 male voices and 12 character voices.

Listen to the samples on the card
Pick a voice and note its file
Load the file in Phonon S
Phonon speaks your text in that voice

Good to know:

• Languages: English, French, German, Spanish and Portuguese
• Voices: 48 female, 42 male and 12 English character voices

Title: Voices per language
English: 37 voices
Spanish: 25 voices
French: 15 voices
German: 13 voices
Portuguese: 12 voices

• License: apache-2.0
• Cost to run: no API key or paid service is named
• Data it reaches: downloading fetches the voice files, and the card names no other access
• Not for: cloning a real person's voice, since these are ready-made voices and use a person's voice only with their right to it

On Hugging Face, 13 November 2026

• Downloads in the last 30 days: 0
• Likes: 39
• License: apache-2.0
• Languages: en, fr, de, es, pt

Links

• Model: https://huggingface.co/gradium/phonon-voices
• Files: https://huggingface.co/gradium/phonon-voices/tree/main

Get the model

Download its files with Hugging Face's command line:

pip install -U huggingface_hub
hf download gradium/phonon-voices

Then follow the usage in its model card. Nothing in this kit runs by itself.