Stonewuu/ai-fusion-video
【融光】 - 基于 Agent 的全流程AI短剧/漫剧/视频创作平台
What it solves
This project provides an integrated platform for content creators to streamline the video production process. It automates the transition from a written script to a finished video by breaking down scripts into visual storyboards and generating the necessary visual assets using AI.
How it works
The platform uses an agent-based approach to handle the video creation pipeline. Users write scripts, which the AI then decomposes into structured storyboards containing scene descriptions and camera language. It then integrates with various AI image and video generation engines to create reference images and video clips based on those descriptions.
Who it’s for
Content creators who want to automate the tedious parts of storyboarding and asset generation for their videos.
Highlights
- Automated Storyboarding: Automatically converts scripts into visual storyboards with scene descriptions.
- Multi-Model Integration: Supports a wide range of LLMs including OpenAI, Claude, Gemini, DeepSeek, and Ollama.
- Asset Management: Centralized management for all generated images and video clips.
- Flexible Storage: Supports local storage as well as S3-compatible services like Aliyun OSS, Tencent COS, and MinIO.
- Agent Pipeline Visualization: Includes a visual flow for the AI agent pipeline.