VoiceTut Egyptian-Arabic TTS (diacritized)

VoiceTut Egyptian-Arabic TTS (diacritized)

Reads Egyptian Arabic text aloud, using diacritics for clearer pronunciation, with 19 built-in voices or a voice cloned from a short clip. It suits creators, learners and developers working in Egyptian Arabic.

Made by hugging-apps 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 turns Egyptian Arabic text into speech. It uses a fine-tuned 0.6B OmniVoice checkpoint trained for text that carries diacritics, called tashkeel, which makes pronunciation clearer. It runs on zero-a10g hardware at Hugging Face and is tagged MCP.

Who it suits: It suits people who need spoken Egyptian Arabic from written text, and developers who want a small Gradio example of a speech model.

Industry | Who uses it | How they use it | Example
Content creation | Video and podcast makers | Turn a script into a voice-over | Read out a short line of Egyptian Arabic with diacritics
Education | Arabic teachers and learners | Hear how a diacritized sentence sounds | Listen to a sentence with its vowels marked
Software | Developers | Study a Gradio text-to-speech app | Read app.py, which is short

You can skip it if you need a tool for any other Arabic dialect, since this one targets Egyptian Arabic.

Where to use it: You can use it in these places:

• A browser: open the Space's page on Hugging Face and nothing needs installing.
• Windows, macOS or Linux: run your own copy with Python 3.12.

Use case | Who | What they get back
Built-in voices | Anyone | Speech in one of 19 reference voices
Voice cloning | Creators with a short clip | Speech in a voice matched to your clip
Pronunciation fixes | Learners and editors | Corrected speech after adding diacritics to a word

How to set it up: You need these first for a local copy:

• Python 3.12 and pip
• A machine that can run the model, which the Space runs in fp16 on a GPU

The online page needs none of this. For your own copy, follow the steps below.

1. Install the Hugging Face tools.
2. Download the Space's files into a folder.
3. Move into that folder.
4. Install Gradio at the version the Space uses, then the other packages.
5. Start the app.

pip install -U huggingface_hub
hf download hugging-apps/voicetut-egyptian-tashkeel --repo-type space --local-dir voicetut-egyptian-tashkeel
cd voicetut-egyptian-tashkeel
pip install gradio==6.31.0
pip install -r requirements.txt
python app.py

How to use it: Open the Space in a browser, or start your own copy with the command below. Then type your text, pick a voice and generate the audio.

python app.py

Tool | What it does
Built-in voices | Offers 19 reference speakers, with a reference clip to listen to before generating
Clone a voice | Takes a reference clip of 3 to 10 seconds plus its exact transcript
Diacritics input | Marks vowels in the text, which improves pronunciation

Example: You type an Egyptian Arabic sentence with its diacritics, pick one of the built-in voices and press generate. The Space returns an audio clip of that sentence in the chosen voice.

For newcomers and experts: If you are new to this, start with a built-in voice and one short sentence. Listen to its reference clip first, then generate and compare.

If you know the field, the model is an OmniVoice fine-tune run through the omnivoice library. It was trained on a diacritized dataset where roughly half of the words got machine-generated vowels, so a bad word can be fixed by adding marks to it. About 2 percent of lines may come out near silent, and the app retries once before it reports an error.

You type diacritized text
You pick or clone a voice
The model generates speech
A silent result is retried once
You play the audio

Good to know: Output is audio at 24 kHz. The voice cloning option needs a reference clip and its exact transcript.

• License: apache-2
• Cost to run: none named
• Data it reaches: the reference clip you upload for cloning, and the model files downloaded from Hugging Face
• Not for: pretending to be a real person, or misleading anyone who listens

On Hugging Face, 13 November 2026

• Likes: 3
• Built with: gradio
• License: apache-2.0

Links

• Space: https://huggingface.co/spaces/hugging-apps/voicetut-egyptian-tashkeel
• Files: https://huggingface.co/spaces/hugging-apps/voicetut-egyptian-tashkeel/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.