dramaclaw  by dramaclaw

A full pipeline for AI-generated video production

Created 3 months ago
754 stars

Top 45.4% on SourcePulse

GitHubView on GitHub
Project Summary

Summary

DramaClaw is a general-purpose AIGC video engine automating the entire production pipeline for content like dramas, ads, and games, from script to film. It targets creators, indie studios, and engineers, aiming to democratize industrialized video production by enabling users to run the full pipeline on their own infrastructure, bypassing opaque cloud services and empowering creative control.

How It Works

The system functions as an industrialized production line, processing manuscripts through stages like character extraction, script generation, storyboarding, and video composition. Its novelty lies in embedding domain-specific "judgment" for narrative elements (e.g., episode beats, character consistency), differentiating it from generic workflow tools. The pipeline is decomposable, allowing independent execution, skipping, or resuming of stages, and features a "dual-track" system combining a main pipeline with an "infinite canvas" for creative exploration.

Quick Start & Requirements

Docker is recommended for installation. Key prerequisites include:

  • System: macOS, Windows (Docker Desktop + WSL2), or Linux (Docker Engine + compose).
  • Resources: ≥ 2 vCPU / 4 GB RAM recommended for local footprint (model inference runs remotely). GPU not required for the standard pipeline.
  • Dependencies: Docker, docker compose. Python 3.11-3.12 and ffmpeg for local development.
  • Network: Outbound access to an OpenAI-compatible model gateway.
  • Links: Docs, [Quick Start](./
Health Check
Last Commit

6 hours ago

Responsiveness

Inactive

Pull Requests (30d)
40
Issues (30d)
42
Star History
745 stars in the last 30 days

Explore Similar Projects

Feedback? Help us improve.