Orange is an intuitive web interface that simplifies ComfyUI, making powerful AI media generation and editing accessible to everyone without the complex node graph.
Python
6
342 commits
updated Oct 2, 2026
Orange is a small web frontend for ComfyUI. It turns finished ComfyUI workflows into simple tools that people can use without seeing the node graph, model setup, or backend details.
Smarter internals, simpler surface.
ComfyUI is where the workflow gets built. Orange is where people use it.
The easiest all-in-one setup is the companion Orange Pinokio launcher.
It offers two install paths:
Pinokio does not download workflow models up front. Orange checks the connected ComfyUI first and only installs missing files for tools you choose.
You need Python 3 and a reachable ComfyUI instance.
Clone the repository, then start Orange:
Windows: run.bat
Linux/macOS: ./run.sh
Open http://localhost:7070/.
The launchers create the Python environment, sync dependencies when needed, keep routine logs quiet, and show useful generation and LLM activity in the terminal. Set ORANGE_VERBOSE_LOGS=1 if you want full live logs.
A fresh install opens a setup wizard. Orange connects to ComfyUI, scans the hardware, nodes, and model inventory, then lets you add curated tools or skip them entirely. Missing model downloads are shown before anything changes, and selected tools go through Workflow Preflight before they are exposed.
You can add curated tools later from Admin → Tools → Curated Library, or import your own API-format ComfyUI workflows.
Orange is released under the MIT License.
Third-party components and model weights keep their own licenses. The curated workflow packs can download models with terms that are more restrictive than Orange itself, so check Third-party notices before using a pack commercially.
Orange should make ComfyUI easier to use without moving workflow-engineering decisions onto the people generating with it. New features should mostly improve onboarding, reliability, diagnostics, and workflow management instead of adding more technical controls to the Generate page.
Python
48.6%
JavaScript
36.0%
HTML
7.8%
CSS
7.2%
Orange is an intuitive web interface that simplifies ComfyUI, making powerful AI media generation and editing accessible to everyone without the complex node graph.
Python
6
342 commits
updated Oct 2, 2026
Orange is a small web frontend for ComfyUI. It turns finished ComfyUI workflows into simple tools that people can use without seeing the node graph, model setup, or backend details.
Smarter internals, simpler surface.
ComfyUI is where the workflow gets built. Orange is where people use it.
The easiest all-in-one setup is the companion Orange Pinokio launcher.
It offers two install paths:
Pinokio does not download workflow models up front. Orange checks the connected ComfyUI first and only installs missing files for tools you choose.
You need Python 3 and a reachable ComfyUI instance.
Clone the repository, then start Orange:
Windows: run.bat
Linux/macOS: ./run.sh
Open http://localhost:7070/.
The launchers create the Python environment, sync dependencies when needed, keep routine logs quiet, and show useful generation and LLM activity in the terminal. Set ORANGE_VERBOSE_LOGS=1 if you want full live logs.
A fresh install opens a setup wizard. Orange connects to ComfyUI, scans the hardware, nodes, and model inventory, then lets you add curated tools or skip them entirely. Missing model downloads are shown before anything changes, and selected tools go through Workflow Preflight before they are exposed.
You can add curated tools later from Admin → Tools → Curated Library, or import your own API-format ComfyUI workflows.
Orange is released under the MIT License.
Third-party components and model weights keep their own licenses. The curated workflow packs can download models with terms that are more restrictive than Orange itself, so check Third-party notices before using a pack commercially.
Orange should make ComfyUI easier to use without moving workflow-engineering decisions onto the people generating with it. New features should mostly improve onboarding, reliability, diagnostics, and workflow management instead of adding more technical controls to the Generate page.
Python
48.6%
JavaScript
36.0%
HTML
7.8%
CSS
7.2%