A modular web UI for AI image, video and audio generation, built on Stable Diffusion, Flux, Wan and other models, with a ComfyUI workflow tab.
akatz-labs/opencast-studio
Open-source app for building AI character reference sheets and using those characters in video via character swaps and motion transfer on ComfyUI.
- Stars
- 37
- Forks
- 7
- Open issues
- 0
- Last commit
- 3 days ago
What is it?
OpenCast Studio is a self-hostable web app for creating character reference sheets and applying those characters to video. It includes a browser frontend, a local API, a GPU executor, ComfyUI workflows, and model manifests. It runs on your own workstation or on a GPU machine reached over SSH.
The project supports character sheet generation, character swaps in existing footage, and motion transfer. On first launch it checks model inventory and runtime status, then guides installation of the bundled ComfyGit toolchain, Krea and Pika environments, and pinned model files. Generation is unavailable until a workflow installation is configured.
Automatic runtime setup currently targets Linux x64 with an NVIDIA GPU. On other platforms, a compatible remote executor is needed. Authentication, billing, and cloud execution are not part of this repository and it is not intended as a public multiuser service.
Key features
- 01Generates character reference sheets from a browser interface.
- 02Swaps characters into existing video footage.
- 03Transfers motion to characters through ComfyUI workflows.
- 04Guided setup checks models and installs pinned runtime dependencies.
- 05Supports local execution or a remote GPU executor over SSH.
- 06Configuration is provided through environment variables.
Use cases
- Creating consistent character designs for short video or animation projects.
- Replacing a performer's character in existing footage on a local GPU workstation.
- Experimenting with motion transfer using ComfyUI workflows.
- Running a private character video pipeline on a GPU machine reached over SSH.
Quick start
- 1
Install dependencies
npm ci - 2
Build the interface and API
npm run build - 3
Start the application and open http://localhost:8389
npm start
Things to know
- Requires Node.js 22.18 or newer and ffprobe for reference-video validation.
- Generation is unavailable until a GPU workflow installation is configured.
- Automatic GPU runtime setup currently targets only Linux x64 with an NVIDIA GPU.
- Gated FLUX3 model weights must be obtained manually after accepting Hugging Face conditions.
In collections
Topics
FAQ
- What is opencast-studio?
- Open-source app for building AI character reference sheets and using those characters in video via character swaps and motion transfer on ComfyUI. OpenCast Studio is a self-hostable web app for creating character reference sheets and applying those characters to video. It includes a browser frontend, a local API, a GPU executor, ComfyUI workflows, and model manifests. It runs on your own workstation or on a GPU machine reached over SSH.
- Is opencast-studio free to use? What license does it have?
- opencast-studio is open source under the MIT license. Check the license terms before using it in commercial products.
- Is opencast-studio still maintained?
- Actively maintained. The last commit to opencast-studio was on Nov 4, 2026.
- What can I use opencast-studio for?
- Creating consistent character designs for short video or animation projects. Replacing a performer's character in existing footage on a local GPU workstation. Experimenting with motion transfer using ComfyUI workflows. Running a private character video pipeline on a GPU machine reached over SSH.
- How do I install opencast-studio?
- Install dependencies: npm ci Build the interface and API: npm run build Start the application and open http://localhost:8389: npm start
- Where can I get opencast-studio?
- The source code and installation instructions are on GitHub: https://github.com/akatz-labs/opencast-studio
Summary generated by AI from the project's README. GitHub stats refresh daily.
Related projects
A modular web UI for Stable Diffusion image generation, built to make advanced tools easy to reach and extensible.
A ComfyUI custom node pack that builds multi-view character reference sheets from a character image using MiniMax H3 image-reference conditioning.
A Gradio-based web UI for running Stable Diffusion and newer image/video models, continuing the Forge fork with added model support and optimizations.
A Nix flake that packages many AI tools, such as ComfyUI, so they can run on NixOS and Linux without manual dependency setup.
A node-graph GUI, API and backend for running diffusion models locally to generate images, video, audio and 3D content.