Podify My Paper: The AI Audio Storyteller

Overview

Podify My Paper instantly transforms your research papers, academic PDFs, or technical documents into engaging, AI-narrated podcast episodes. No technical skills required! Just upload your PDF, choose your preferred podcast style, language and let advanced AI models craft a natural, insightful dialogue between two expert speakers. Download and share your audio podcast in minutes.

Whether you want a deep-dive discussion or a concise highlights summary, Podify My Paper brings your ideas to life through conversational storytelling.

What’s New

• Language Selection:
  Supports English, Hindi, German, and French. Generate podcasts in your preferred language.
• Multi-Voice, Multi-Language Audio:
  Murf AI produces high-quality voices for both "Nishant" and "Megha" in all supported languages (where available).

Demo Video

See Podify My Paper in action!

demo.mp4

Flow Diagram

graph TD
    A[User uploads PDF] --> B[Extract & Chunk Text]
    B --> C{Podcast Mode?}
    C -- Full --> D[Generate Full Script]
    C -- Summary --> E[Summarize, then Script]
    D --> F[Text-to-Speech for Each Chunk - Selected Language & Voices]
    E --> F
    F --> G[Combine Audio to MP3]
    G --> H[Download Podcast]
    H --> I[Clean Up Temporary Files]

Loading

How It Works

1. Upload PDF
   Upload any academic, research, or technical PDF using the web interface.

2. Choose Podcast Style

   • Full Podcast (All Details): Converts your entire document into a detailed, conversational podcast.
   • Highlights Podcast (Summary Only): Generates a concise, summary-focused podcast episode.

3. Select Language
   Choose from English, Hindi, German, or French for your podcast’s narration and script.

4. AI Script Generation

   • Uses Google Gemini (Generative AI) to analyze your document, summarize content (if needed), and create a natural back-and-forth script between "Nishant" (host) and "Megha" (expert guest).
   • For longer documents, splits content into chunks to ensure smooth, manageable podcast segments.

5. Text-to-Speech Audio Creation
   Each line of the script is converted to speech using realistic voices. You can now choose between:

   • Murf AI TTS (multi-voice, cross-platform):
     The default, recommended option, with natural-sounding voices for both host and guest. No platform restrictions!

   • macOS Say Command (optional):
     For Mac users who want to use built-in lifelike voices.

   • gTTS (optional):
     For simple, single-voice cross-platform support.

Note: Murf AI is now integrated and used by default for the best-quality, multi-speaker experience.

6. Download Podcast

   • All audio segments are combined into a single MP3.
   • Download and listen to your personalized podcast episode!

7. Automatic Cleanup
   Temporary audio chunks are deleted after the MP3 is created.

Features

• PDF-to-Podcast: Converts any research or technical PDF into a two-speaker podcast.

• Multi-Language Support: Supports English, Hindi, German, and French with AI-powered translation.

• Multi-Voice Support: Both host and guest have their own unique, realistic AI voices (via Murf).

• AI Summarization: Generate either a detailed or summary-based podcast episode.

• Web UI: Easy-to-use Gradio interface—no coding required.

• Automatic Cleanup: Cleans up all intermediate audio files from podcast_chunks after each run.

Requirements

• Python: 3.8 or higher
• Google Gemini API Key: Required for AI podcast script and summarization. Sign up here.
• Murf AI API Key: Required for Murf TTS voice generation. Get your Murf API key here.
• Python Dependencies: See requirements.txt.

Installation

git clone https://github.com/nishanttomar21/podify-my-paper.git
cd podify-my-paper
pip install -r requirements.txt

Usage

1. Set up API Key(s):

You'll need API keys for:

• Google Gemini (Generative AI)
• Murf AI TTS

You can provide your API keys in one of two ways:

Option 1: Directly edit the code (main.py) and replace:

GEMINI_API_KEY = "Enter-your-api-key"
MURF_API_KEY = "Enter-your-murf-api-key"

Option 2 (Recommended): Set the keys as environment variables in your terminal:

export GEMINI_API_KEY=your_actual_gemini_key_here
export MURF_API_KEY=your_actual_murf_key_here

On Windows, use:

set GEMINI_API_KEY=your_actual_key_here
set MURF_API_KEY=your_actual_murf_key_here

2. Run the App

After installation and setting the API key, start the web app with:

python main.py

Changelog

v2.0

• Language selection: English, Hindi, German, French
• Translation: Scripts are translated automatically, and multi-voice TTS is supported per language
• Automatic cleanup: The podcast_chunks folder is cleaned after MP3 creation

Credits

• Gradio – Easy-to-use web UI for the app
• Google Gemini API – AI script generation & summarization
• Murf AI – High-quality, multi-voice text-to-speech
• gTTS – Alternative simple TTS engine
• pydub – Audio processing and MP3 export

License

MIT License