What is OpenDirector?

OpenDirector is an open-source AI video production platform where 9 specialized AI agents collaborate to produce complete videos from a single sentence prompt. Built on LangGraph for orchestration, Next.js for the frontend, and FFCreator for rendering, it automates the entire pipeline — from scriptwriting and storyboarding to voiceover, background music, and final video render. It supports both interactive creative mode and automated batch production.

How does OpenDirector work?

OpenDirector uses a 9-agent pipeline orchestrated by LangGraph. The Research Agent gathers context, the Script Agent writes the narrative, the Art Style Agent selects from 34 visual styles, the Storyboard Agent plans shots, Character and Location Agents design visuals, the Voice Agent generates voiceovers, the BGM Agent composes music, and the Media Agent renders the final video. Each agent communicates through a state graph, enabling real-time collaboration and manual adjustments at any stage.

Is OpenDirector free?

Yes, OpenDirector is completely free and open-source under the MIT license. You can use it for personal and commercial projects. The only costs are for the AI API keys you configure (such as OpenAI, Anthropic, or other LLM providers) and any TTS or image generation services you connect.

How do I deploy OpenDirector?

OpenDirector is Docker-first. Clone the GitHub repository, configure your API credentials in the .env file, then run docker compose up to boot MySQL, Redis, MinIO, and the app. Open your browser to start creating. The entire setup takes about 5 minutes on any machine with Docker installed.

What video formats and aspect ratios does OpenDirector support?

OpenDirector supports three aspect ratios: 16:9 cinematic widescreen, 9:16 vertical portrait, and 1:1 square format. Videos can be exported at up to 1080p Full HD resolution. The platform includes 34 built-in art styles across 9 categories, from cinematic and anime to watercolor and neon noir.

Is my data private with OpenDirector?

Yes. Since OpenDirector is self-hosted via Docker, all your scripts, voiceovers, character designs, and rendered videos stay on your local machine. Nothing is uploaded to third-party servers. The only external calls are to the AI APIs you explicitly configure for generation tasks.

Deploy OpenDirector | Quick Start & Installation

Self-Host with One Command

OpenDirector is Docker-first. Keep your transcripts, files, and assets completely private on your local machine.

Clone repository

Terminal step 1

Configure API credentials in env file

Terminal step 2

Boot up MySQL, Redis, MinIO, and App

Terminal step 3

Open in browser to start creating

Terminal step 4

self_host_setup.sh

git clone https://github.com/seme-org/open-director.git
cd open-director

cp .env.example .env
# Edit .env and enter your WAVESPEED_API_KEY and OPENAI_API_KEY

docker compose up -d --build

# Open http://localhost:3000 in your browser

Prerequisites

🐳

Docker

Docker Desktop 4.0+ or Docker Engine 20.10+

💻

System Requirements

4GB+ RAM, 10GB+ disk space

🔑

API Keys

At least one LLM API key (OpenAI, Anthropic, etc.)

Advanced Configuration

Environment Variables

Configure these key variables in your .env file:

OPENAI_API_KEY=sk-...

ANTHROPIC_API_KEY=sk-ant-...

MYSQL_HOST=mysql

REDIS_HOST=redis

MINIO_ENDPOINT=minio:9000

Service Components

Docker Compose automatically starts: MySQL (data storage), Redis (cache and queue), MinIO (file storage), and the OpenDirector app. All data is persisted in Docker volumes.

Troubleshooting

Docker Compose fails to start

Ensure Docker Desktop is running and ports 3306, 6379, 9000 are not in use. Run docker compose down -v to clean up and retry.

API key errors

Check that API keys in .env are correct. Ensure keys have sufficient credits and permissions.

Video rendering fails

Ensure sufficient system memory (8GB+ recommended). Check that MinIO service is running.

Explore Features →

Learn about OpenDirector's dual production modes

Learn the Pipeline →

See how the 9 AI agents collaborate