This is a web app made with Python, Flask, MoviePy and OpenAI that transforms an audio storytelling file into an enchanting video. Simply upload any audio narration, and this tool will:
- 🎙️ Transcribe your audio using the advanced speech recognition provided by OpenAI Whisper.
- ✂️ Split the transcript into multiple scenes, ensuring each segment of the story is neatly captured.
- 🖌️ Generate dynamic, illustrative images for each scene based on intelligent prompts and style guidelines.
- 🎞️ Compose a video by synchronizing the generated images with the original audio, crafting a cinematic outcome.
- 💫 Automatic Story Extraction: Detects characters, scenarios, and important items from your transcript, giving you a structured "story ingredients" overview.
- 🎨 Whimsical Image Generation: Transforms each scene into a painterly, whimsical illustration that reflects the heart of your story.
- 🪄 One-Click Video Assembly: Seamlessly merges the generated visuals with your narration into a final video, ready to play or share.
- 🌱 Friendly Web Interface: Upload audio, preview and regenerate images if needed, and watch your story bloom into a mini cinematic production.
Whether you’re narrating a fairy tale, sharing personal anecdotes, or preparing a memorable presentation, this tool helps you transform words into visuals—so every story can shine!
- Python 3.7+
- ffmpeg installed on your system
- The packages in
requirements.txt .envfile with yourOPENAI_API_KEY
Create a virtual environment (e.g., named venv):
python -m venv venvOn Windows:
venv\Scripts\activateOn macOS/Linux:
source venv/bin/activateInstall the required packages:
pip install -r requirements.txtCreate a .env file in the root directory and add your OpenAI API key:
OPENAI_API_KEY=your_openai_api_keyRun the script:
python create_video.py ./demo/audio-file.mp3- Integrate the new GPT-4o images API as soon as it comes out.
- Allow editing generated images and provide previous scene images for context, waiting on gpt-4o release to see options.
- Implement a smooth and subtle Kenburns random walk to add some dynamism.
APR 22, 2025:Added feature to select local image and image cropping.APR 21, 2025:Added fade-in, fade-out and cross-fade transitions.APR 18, 2025:Liked the idea and added a UI for more control of each scene.APR 17, 2025:Had the idea and started the project as a simple script.