Skip to main content

​
Parallel Edges (Default)

Purpose: Standard execution flow between nodes Parallel edges are the default connection type. When a node completes, execution flows to all connected nodes simultaneously. Parallel Edges - Start node connected to multiple nodes running concurrently In this example:
  • The Start Node has one PARALLEL output handle
  • A parallel edge connects to the Agent Node
  • The Agent Node executes when Start completes
  • The Agent receives the Start node’s output data
Characteristics:
  • Concurrent Execution: All target nodes run in parallel
  • No Waiting: Each node starts as soon as its source completes
  • Data Cloning: Source output is copied to all targets
Use Cases:
  • Parallel API calls
  • Multi-path workflows
  • Fan-out patterns
  • Independent processing streams

​
What’s Next?

Tool Edges

Connect LLMs to tools for AI function calling

Conditional Edges

Route execution based on conditions

Error Edges

Handle node execution failures

All Edges

View all edge types overview