GeneraVoice

GeneraVoice is a free, Hugging-Face–powered Streamlit app that turns any research abstract into a AI-generated narration with downloadable audio

Turn any research abstract into a narrated explainer.
Built with Hugging Face (FLAN-T5 + SpeechT5) and Streamlit — all free, no pro/paid voices.

Features

• Generative text → concise narration from your abstract (google/flan-t5-base)
• Generative audio → speech from narration (microsoft/speecht5_tts + microsoft/speecht5_hifigan)
• Simple UI → paste, click, listen, download WAV

Quick Start

git clone https://github.com/AnisulMahmud/GeneraVoice.git
cd GeneraVoice
python -m venv .venv
# Windows PowerShell
.\.venv\Scripts\Activate.ps1

pip install -r requirements.txt
streamlit run app.py

Note: This project uses open, non-premium. Voice quality and pronunciation may vary, and speaking rate can be faster than human speech. That’s expected for free models.