---
title: Speak Fongbe with Meta’s Multilingual TTS Model
emoji: 🗣️
colorFrom: pink
colorTo: red
sdk: gradio
sdk_version: "5.38.2"
app_file: app.py
pinned: false
---

# 🔊 Speak Fongbe with Meta's Multilingual TTS Model

## 🗣️ Fongbe Text-to-Speech (TTS) Demo  
**Model:** [`facebook/mms-tts-fon`](https://huggingface.co/facebook/mms-tts-fon)  
**Architecture:** VITS (Vocoder-based TTS)  
**Language:** Fon (ISO 639-3: `fon`)  
**License:** CC-BY-NC 4.0  
**Powered by:** Meta’s Massively Multilingual Speech (MMS) project

---

## 📖 Description (Fongbe)  

**Gbè elɔ xlɛ́ kpɔ́ndéwú nǔwlanwlán tɔn e è sɔ́ ɖ’ayǐ é ɖé nú Fongbe**, gbè taji e è nɔ dó ɖò Benin kpo Togo sín akpáxwé ɖé lɛ kpo mɛ é.  
**Kpɔ́ndéwú:** facebook/mms-tts-fon  
**Xɔgbigbá:** VITS (TTS e jinjɔn Vocoder jí é)  
**Gbè:** Fɔn (ISO 639-3: fon)  
**Acɛ:** CC-ƉO-NC 4.0  
**Powered by:** [Mɛta sín xóɖiɖɔ gbè gègě tɔn (MMS)](https://huggingface.co/facebook/mms-tts-fon)

---

### 🔍 Lee è nɔ zán gbɔn é  
Wlan xógbe klewun ɖé dó **Fon** mɛ bo zín “Submit” bo na dó sè gbè e è xò kplé é. Modèle ɔ nɔ ɖè xóɖiɖɔ tɔ́n ɖò hwenu e é jɛ é, bo nɔ zán vocoder gbè gègě tɔn e è ko kplɔ́n ɖ’ayǐ é ɖé.

---

### 🌍 Gbè ɔ wú  
Fon ɔ, Niger-Congo gbè wɛ bɔ gbɛtɔ́ livi 2 jɛji wɛ nɔ dó. È nɔ zán ɖò wemaxɔmɛ, xójlajla sín nǔ lɛ kpo xóɖɔɖókpɔ́ ayihɔngbe ayihɔngbe tɔn lɛ kpo mɛ, amɔ̌, è kpó ɖò xóɖiɖɔ sín nǔnywɛ xwitixwiti sín nǔ lɛ zán wɛ ganji ǎ.  
Demo enɛ ɔ nɔ d’alɔ bɔ è nɔ sú dò enɛ gbɔn TTS ɖiɖó bɔ é nɔ bɔkun nú gbè Aflika tɔn e mɛ nǔɖokan lɛ ma sukpɔ́ ɖè ǎ lɛ é gblamɛ.

---

## 📖 Description (English)

This Space showcases a cutting-edge text-to-speech model for **Fongbe**, a major language spoken in Benin and parts of Togo.  
**Model:** [`facebook/mms-tts-fon`](https://huggingface.co/facebook/mms-tts-fon)  
**Architecture:** VITS (Vocoder-based TTS)  
**Language:** Fon (ISO 639-3: `fon`)  
**License:** CC-BY-NC 4.0  
**Powered by:** Meta’s Massively Multilingual Speech (MMS) project

---

### 🔍 How to Use  
Type a short sentence in **Fon** and press “Submit” to hear the synthesized voice.  
The model outputs audio in real-time using a pretrained multilingual vocoder.

---

### 🌍 About the Language  
Fon is a Niger-Congo language with over 2 million speakers.  
It is used in education, media, and daily communication, yet remains underrepresented in speech technology.  
This demo helps close that gap by making TTS more accessible for low-resource African languages.

---

🎤 Built with 🤗 Gradio & Meta MMS  
👤 Demo created by [Bonaventure F. P. Dossou](https://bonaventuredossou.github.io/)