Anil-matcha/Open-AI-Micro-Drama-Generator

AI short drama & micro-drama video generator — turns any idea into a complete short-form drama using multi-agent AI pipeline (screenwriter → storyboard → frames → video). Seedance 2 VIP, Kling 3.0 Pro, Veo 3.1, Sora 2.

What it solves

MicroDrama AI addresses the fragmented nature of AI video production, where users typically have to manually switch between separate tools for scripting, image generation, and video creation. It solves the problem of narrative incoherence and character inconsistency by automating the entire pipeline from a simple idea to a final cinematic video.

How it works

The system uses an agentic pipeline powered by MuAPI to orchestrate several specialized AI roles:

  1. Screenwriter Agent: Develops the story outline and writes scene scripts.
  2. Character Extractor Agent: Identifies characters and defines their static and dynamic visual features.
  3. Storyboard Artist Agent: Designs a shot-by-shot breakdown including visual, motion, and audio descriptions.
  4. Production Tools:
    • Generates reference portraits using flux-dev-image (T2I).
    • Creates first frames for shots using flux-kontext-dev-i2i (I2I) to maintain character consistency.
    • Produces 5-second video clips using kling-v2.1-standard-i2v (I2V).
  5. Concatenation: Uses moviepy to stitch all generated clips into a final .mp4 file.

Who it’s for

It is designed for creators who want to transform raw ideas or existing scripts into short cinematic videos without managing multiple AI tools or manually ensuring visual continuity between shots.

Highlights

  • End-to-End Automation: Converts a raw idea or a provided script into a full video with zero manual steps.
  • Character Consistency: Uses reference portraits and I2I generation to keep characters looking the same across different shots.
  • Real-time Monitoring: Features a Next.js frontend with SSE streaming to show live progress through a 7-stage pipeline.
  • Single API Integration: Operates entirely using a single MuAPI key for LLM, image, and video generation.