Nodes are the fundamental components that power your AI workflow system. Think of them as specialized building blocks that perform specific functions and work together to create complex processes. This guide will help you understand what nodes are, how they connect, and the basics of working with them.
What Are Nodes?
A node is a self‑contained action card that represent specific AI tools or functions within your workflow. Each time you press Generate, the AI creates or updates that card, producing a defined output.
Types of Nodes
Text Model Nodes: Process and generate text using various AI models like GPT-4o, Claude, Gemini, and others
Image Nodes: Create or edit images using tools like DALL-E 3, Flux, Stable Diffusion, or Ideogram
Video Nodes: Turn an idea or an image into a short video animation with Kling AI
Research Nodes: Gather information from the web using tools like Perplexity, Google Search, or Deep Research
Scraper Nodes: Extract data from websites or social media platforms
Utility Nodes: Help organize your workflow with elements like sticky notes, groups or text blocks
How Nodes Work
Each node:
Takes inputs from preceding nodes or direct user input
Processes that information using its specialized AI capability
Produces outputs that can be used independently or fed to subsequent nodes
When you add a node to your workspace, you're essentially adding a new AI capability to your workflow.
Connection Points
Nodes communicate with each other through connection points, allowing information to flow through your workflow.
Understanding Connection Points
Input Points (left side of nodes): Accept incoming data from other nodes
Most AI nodes have one input point for receiving prompts
Some image nodes have a second input point for receiving reference images
Output Points (right side of nodes): Send processed data to other nodes
Understanding Data Flow
Information flows from left to right across your whiteboard
Output from one node becomes input for connected nodes
This creates a pipeline where data is progressively transformed
This chaining of nodes creates a workflow where each node builds upon the work of previous nodes. Creating thoughtful connections between nodes allows for complex, multi-step AI operations
For example, you could create a flow where:
A Research node gathers information about a topic
The research results flow into a Text node that summarizes the findings
The summary flows into an Image node that creates a visual representation
Basic Node Operations
Here's how to perform essential operations with nodes:
Adding Nodes
Go to your toolbar and select the category of node you want (Text, Image, Research, etc.)
Choose the specific AI tool you want to use and click in the '+' button to add to your whiteboard
Configuring Nodes
Click the ‘⚙ Configure' button on any node to open its settings panel.
Set parameters specific to that node type:
For Text nodes: Select the model and set the prompt.
For Image nodes: Enter your prompt, choose image size, quality and style settings.
For Research nodes: Define your research query, depth, and topic.
E.g.
Note: Each Node configuration mode may vary depending on the specific setup of its underlying AI Agent.
Connecting Nodes
Click on any point of a source node
Drag the line to the target node to create a connection
A visual line will appear to indicate the connection
The receiving node will turn white to show that a prompt or a contextual information has already been added.
Executing Nodes
After configuring your nodes and establishing connections:
When a node is ready to run, a 'Generate' button will appear at the top right corner—click it to process that specific step.
After a few seconds, the final result will appear within the node.
Deleting Nodes
Select the node you want to remove
Press the Delete key, or right-click and select "Remove"
Deleting Connections
Zoom in on the area where the connection line is located.
Click the connection line and wait for it to highlight (the color will become more intense to indicate it’s selected).
Press Delete on your keyboard to remove the connection.
By understanding these node fundamentals, you'll be able to create increasingly sophisticated AI workflows that automate complex tasks and enhance your productivity.