Story-Flicks: one-click AI-generated story videos

Making a story-style short video usually means bouncing between several tools for the script, illustrations, narration, and subtitles. Story-Flicks folds that whole pipeline into a single open-source project: give it a story theme, and it has an LLM write the story, illustrates each section with AI-generated art, adds narrated audio and matching subtitles, and finally composites it all into one complete video. GitHubDaily recommended this project in March 2025.

Core features

• One-click video: give it just a theme, and the script, illustrations, narration, and subtitles are all generated automatically, ending in a finished video.
• Multiple model providers: for text, choose DeepSeek or OpenAI, or connect to Alibaba Cloud, a locally deployed Ollama, or SiliconFlow; for images, OpenAI, Alibaba Cloud, and SiliconFlow are supported—just fill in the base_url and API key for each to switch providers.
• Web UI operation: a FastAPI backend and React frontend (Ant Design + Vite) let you pick the provider, model, language, voice, theme, and number of sections right in the page, click generate, and see the finished video displayed there.
• Multiple languages and voices: both the video’s language and the voice used for narration are configurable, making it easy to produce content for different audiences.
• Customizable sections: the number of story sections is adjustable, with one image generated per section—more sections mean a longer wait.

Typical use cases

• Children’s bedtime stories and fable-style short videos: the project’s own demos use themes like “the rabbit and the little fox” and “the big bad wolf and the little white rabbit.”
• Ongoing production of history, science-explainer, and similar story content for short-video platforms.
• Developers building on its FastAPI + React architecture to plug in their own model services and extend the video-generation pipeline.

Quick start

1. Clone the repo and configure your keys:

git clone https://github.com/alecm20/story-flicks.git
cd backend && cp .env.example .env
# Edit .env to set text_provider, image_provider, and the corresponding API keys

2. Start everything with one command from the project root:

docker-compose up --build

You can also install manually: run the backend with Python 3.10 (a conda environment) plus uvicorn, and for the frontend run npm install && npm run dev, then open http://localhost:5173/.

Summary

This suits creators making children’s stories, fables, or science-explainer short videos, as well as developers who want to quickly stand up an AI video-generation pipeline or plug in their own model. Overall it’s a tool-style project that runs as soon as you’ve configured an API key, with a low barrier to entry. A few caveats: the repo currently has no LICENSE file, so confirm the terms before reusing or commercializing it; choosing DeepSeek or Ollama gives you text generation only, with no image output; if using Ollama locally, a model of 14B parameters or larger is recommended; and the project’s last code push was in March 2025 with no updates since, so maintenance activity is low—evaluate it carefully before adopting it.