When you open a project in FlowScale AI, you enter the ComfyUI Workspace — your complete environment for building, testing, and managing generative AI workflows. Think of it as your personal AI laboratory where the familiar ComfyUI interface meets enterprise-grade cloud infrastructure.
The workspace seamlessly bridges the gap between creative experimentation and production deployment. You start with an idea, build it using ComfyUI’s visual workflow editor, test it with your team’s shared resources, and then deploy it to production — all within a single, integrated environment.
Your workspace is thoughtfully organized into three key areas, each serving a specific purpose in your workflow development journey:Left Panel: Your Project Hub
This is where you manage everything related to your project. The Workflows tab helps you organize your creations, the Models tab gives you access to your organization’s AI model library, and the Nodes tab manages the custom functionality your workflows depend on.Right Panel: Your Creative Canvas
Here’s where the magic happens. After clicking “Start ComfyUI”, this area transforms into the full ComfyUI interface where you’ll design and test your AI workflows.Bottom Panel: Your Control Center
Think of this as your mission control — buttons to save your work, toggle between CPU and GPU modes, run your workflows, and monitor everything through the logs.
Click “Start ComfyUI” in the right panel — this is your gateway to the creative process
Choose your computing power by selecting a Pod from the dialog
Watch the magic happen as the system prepares your environment (you’ll see “ComfyUI resources are being prepared”)
Enter your creative space as ComfyUI loads with a default workflow ready for your modifications
Once ComfyUI springs to life, you’re greeted with a pre-loaded workflow that serves as your starting point. The full ComfyUI interface is now at your fingertips, enhanced with FlowScale’s collaborative features and cloud infrastructure.
Now that ComfyUI is running, let’s explore the control bar at the bottom — your command center for managing workflows efficiently and cost-effectively.
Each button serves a specific purpose in your creative workflow:
Here’s where FlowScale’s cost-effectiveness shines. Start in CPU mode ($0.14/hour) for workflow building and editing, then switch to GPU mode only when you need to run actual generations. This approach can save your organization significant costs while maintaining full functionality.
Ready to see your workflow in action? This button executes your current workflow, processing it through the entire pipeline with your specified inputs.
When you’re done working, this properly shuts down your ComfyUI session and frees up computing resources. You’ll need to restart ComfyUI when you’re ready to work again.
The Workflows tab is where your AI creations come to life and get organized. Whether you’re starting from scratch or bringing existing workflows into FlowScale, this is your creative command center.
Organize with Purpose
Click the plus icon to create folders that make sense for your team — organize by project type, client, or development stage. A well-structured workspace saves time and reduces confusion as your library grows.Begin Creating
The “Create New Workflow” button loads a default ComfyUI template, giving you a clean slate to build upon. It’s like having a prepared canvas waiting for your artistic vision.
Simply drag and drop your workflow JSON file (use the standard ComfyUI workflow format, not the API version), give it a descriptive name, and choose where to store it.
As your workflow collection expands, the three-dot menu on each item gives you essential management options: rename for clarity, download for backup, or delete when no longer needed. The visual indicators (red cross or green checkmark) instantly show you which workflows are ready for API deployment.
The Models tab brings your organization’s entire AI model collection directly into your workspace. It’s the same powerful interface you’d find in the global Models section, but conveniently located where you’re actively building workflows.This means while you’re deep in creative flow, you can quickly browse, add, or manage the AI models your workflows depend on — without leaving your workspace. For detailed model management guidance, see our Models documentation.
Think of custom nodes as specialized tools that expand what your workflows can accomplish. The Nodes tab manages these extensions, ensuring your workflows have all the functionality they need.
When you import workflows, any required custom nodes appear here automatically. You can see what’s installed, what’s needed, and the status of each dependency at a glance.
Sometimes you’ll want to add custom nodes for future projects or specific functionality:
Click the plus button and select “Add Custom Node”
Search by name for popular nodes, or provide a GitHub repository URL for specific ones
Configure advanced settings if needed:
The advanced settings let you specify Python dependencies, Linux packages, or pin to specific GitHub commits for production stability. Most of the time, the basic settings are sufficient, but these options give you fine-grained control when needed.
Ready to begin your AI workflow adventure? Here’s your roadmap:
Open your project in FlowScale AI and familiarize yourself with the workspace layout
Start ComfyUI by selecting a pod that matches your current needs
Create your first workflow or import an existing one to get hands-on experience
Install any required custom nodes to ensure everything works smoothly
Test your workflow with sample inputs to verify everything functions correctly
Save your work regularly and establish good organizational habits from the start
The ComfyUI Workspace is designed to grow with you — from your first simple workflow to complex production deployments. It bridges the gap between creative experimentation and enterprise-grade deployment, giving you the best of both worlds.