gitmyhub

huobao-drama

TypeScript ★ 13k updated 2d ago

🎬 火宝短剧 - 基于AI的一站式短剧生成平台 《一句话生成完整短剧,从剧本到成片全自动化》 Huobao Drama - An AI-Powered End-to-End Short Drama Generator "One Sentence to Complete Drama: Fully Automated from Script to Final Video"

An AI platform that turns a one-sentence idea into a complete short-form video drama by automatically generating the script, characters, visuals, voiceovers, and final assembled video.

TypeScriptNuxt 3Vue 3HonoSQLiteFFmpegDockersetup: hardcomplexity 4/5

Huobao Drama is an AI-powered platform that turns a one-sentence idea into a complete short-form video drama. You describe what you want, and the system generates the script, designs the characters, creates the visual scenes, adds voiceovers, and assembles the final video, all without you touching any of those steps individually.

The process works through a pipeline of five AI agents. One rewrites your input into a properly formatted screenplay. A second extracts the characters and scenes. A third breaks the screenplay into individual shots. A fourth assigns a voice to each character. A fifth generates the image prompts needed to produce the visuals. The platform supports multiple AI service providers for different tasks, including OpenAI, Gemini, MiniMax, and Alibaba for image generation, and several providers for video clips and text-to-speech.

The technical stack is split between a Nuxt 3 and Vue 3 frontend and a backend built on the Hono web framework. Data is stored in a local SQLite database that the application creates automatically on first start. Video rendering uses FFmpeg, which must be installed separately (or is included if you run the Docker version). Configuration of which AI service providers to use and their API keys happens through a YAML file and the web interface rather than environment variables.

Deployment can be done with Docker Compose, which packages the frontend and backend into a single image with FFmpeg already included. A traditional server setup is also documented, with an example Nginx reverse proxy configuration. The project is released under a Creative Commons BY-NC-SA 4.0 license, meaning it is free for non-commercial use with attribution.

The project is primarily documented in Chinese, though the core concepts and setup steps follow a standard web application pattern and can be followed with the help of a translation tool.

Where it fits