Dafthunkdafthunk

Nodes Reference

In Dafthunk, Nodes are the fundamental units of any workflow. Each node represents a specific operation, function, or action. By connecting these nodes, you visually design the logic and data flow of your automated processes. They are the essential components that make visual programming in Dafthunk powerful and intuitive.

Explore Nodes

The Nodes catalog on this site is a browsable reference of every available node, organized by category. Once you're signed in, the Playground in the app lets you explore the same library interactively — search for specific functionality and see detailed information about each node's inputs, outputs, and parameters. You can test nodes directly with sample data before adding them to your workflows.

Anatomy of a Node

Each node in Dafthunk, regardless of its specific function, shares a common structure designed for clarity and ease of use. Understanding this anatomy is key to effectively building workflows:

  • Inputs: Connection points where data or control signals enter the node. Nodes can have multiple inputs, each expecting a specific type of data.
  • Outputs: Connection points from which data or signals exit the node after processing. The results of a node's operation are passed to other nodes through its outputs.
  • Parameters: Configurable settings that allow you to customize the behavior of a node. Parameters might include API keys, specific values for operations, or choices that dictate how the node functions.
  • Tag: Nodes are grouped into tags (e.g., AI & ML, Text, Data) based on their primary function, making them easier to find and understand.
  • Compatibility: Information detailing which types of workflows or environments support the specific node.

Node Categories

To help you navigate the extensive library, nodes are organized into logical categories based on their core functionality:

  • AI & ML: Leverage powerful language models, image generation capabilities, and other machine learning tools.
  • Text: Perform a wide range of text processing tasks, including formatting, extraction, and manipulation.
  • Image: Tools for image generation, analysis, conversion, and transformation.
  • Audio: Nodes for speech processing, audio synthesis, and other sound-related manipulations.
  • Network: Make HTTP requests, scrape web content, and integrate with various external APIs and services.
  • Data: Process, transform, and manipulate data structures like JSON, XML, and CSV.
  • Math: Perform mathematical operations, calculations, and logical comparisons.
  • Parameters: Specialized nodes for handling inputs, form data, and workflow arguments.
  • Email: Send and process emails, enabling automation of communication tasks.
  • Document: Tools for processing, converting, and extracting data from various document formats.

Connecting Nodes

Workflows come to life by connecting nodes. This process defines how data flows and the sequence of operations:

  1. Establish Connections: Drag from an output port of one node to an input port of another. This creates a visual link representing the path of data.
  2. Data Transmission: Once connected, data produced by an output port automatically flows to the connected input port of the subsequent node.
  3. Type Validation: The system often performs type validation to ensure that the data type of an output is compatible with the expected input type, preventing errors.
  4. Execution Order: The connections form a directed graph, which dictates the order in which nodes execute. Dafthunk processes nodes based on these dependencies.