openmontage in docker ?

[dardizzola]

is it possible to install openmontage in docker ?

[calesthio]

Yes, it should be possible to run OpenMontage in Docker, but we don’t currently provide an official Docker image or Docker Compose setup.
You’d need a container with Python, Node.js, FFmpeg, and the project dependencies. Your AI coding assistant would also need access to the container to run commands, and you’d mount a folder to keep your generated videos.
Are you planning to use cloud AI providers, or run models locally on a GPU? That will determine the setup you need.

[aimanmalib]

Yes. Here is a Dockerfile I tested on Linux: the image builds and the zero-key world-in-numbers demo renders inside the container as a non-root user (1920x1080 H.264 with AAC audio, 23 s). The other two demos rendered on earlier builds of the same file.

FROM python:3.12-slim-bookworm

# FFmpeg for assembly, Node 22 for Remotion and HyperFrames (HyperFrames needs >= 22),
# and the shared libraries the headless Chrome used by Remotion links against.
RUN apt-get update && apt-get install -y --no-install-recommends \
        ffmpeg curl ca-certificates gnupg git make \
        libnss3 libatk1.0-0 libatk-bridge2.0-0 libcups2 libdrm2 libxkbcommon0 libxcomposite1 \
        libxdamage1 libxfixes3 libxrandr2 libgbm1 libasound2 libpango-1.0-0 libcairo2 fonts-liberation \
    && curl -fsSL https://deb.nodesource.com/setup_22.x | bash - \
    && apt-get install -y --no-install-recommends nodejs \
    && rm -rf /var/lib/apt/lists/*

WORKDIR /app
COPY requirements.txt .
RUN pip install --no-cache-dir -r requirements.txt && pip install --no-cache-dir piper-tts
COPY remotion-composer/package*.json remotion-composer/
RUN cd remotion-composer && npm install
COPY . .
RUN cp -n .env.example .env

# Run as a regular user (uid 1000) so renders on a mounted folder are not owned by root.
# Remotion and webpack write caches under node_modules/.remotion and .cache, so those must be writable too.
RUN useradd -m -u 1000 montage \
    && mkdir -p remotion-composer/node_modules/.remotion remotion-composer/node_modules/.cache projects \
    && chown -R montage:montage remotion-composer/node_modules/.remotion remotion-composer/node_modules/.cache projects
USER montage

# Productions and renders are written under /app/projects; mount it to keep them.
VOLUME ["/app/projects"]
CMD ["python", "render_demo.py", "--list"]

Put a .dockerignore next to it. Leaving out .env keeps your API keys out of the image:

.git
.env
node_modules
.venv
output
projects

Build and render a demo from the repo root:

docker build -t openmontage .
mkdir -p projects
docker run --rm --env-file .env -v "$PWD/projects:/app/projects" openmontage python render_demo.py code-to-screen

The video lands in projects/demos/renders/code-to-screen.mp4. python render_demo.py --list shows the other demos.

Notes:

• Create projects before the first run. If Docker creates the folder for you, it belongs to root and the container user cannot write to it.
• The container runs as uid 1000, the usual first user on Linux. If your uid is different, change -u 1000 in the Dockerfile.
• --env-file .env passes your provider keys at run time. Without keys the demos still render, as shown above.
• The image is about 2.6 GB, mostly Chrome libraries, Node modules and the Python packages.
• This is the CPU path. For local GPU models you would start from a CUDA base image and install requirements-gpu.txt, which I have not tested.
• Your AI coding assistant still runs on your machine; point it at the container with docker exec, or run the assistant inside the container.