Aller au contenu principal

Data Splitter Node

The Data Splitter Node allows you to break down long text data into smaller, manageable parts. You can split text using AI-based suggestions or custom separators, making it easier to process structured or semi-structured data in your workflow.

AI Mode​

In AI Mode, the node leverages GPT (default: GPT-4o) to intelligently identify and separate distinct ideas or concepts within the input text.

  • The AI model follows a predefined prompt instructing it to split ideas using semicolons (;).
  • Example use case: If you input a long description or a bullet-point list, the AI will return structured, distinct parts separated by semicolons.
  • The final output is automatically split into a list based on the semicolon separator.

Separator Example

When to Use AI Mode​

  • Works well for short prompts or text with clearly distinguishable elements.
  • Not recommended for long or complex text, as AI may not always split content reliably.

In Separator Mode, you can define a custom separator (default: ;) to split the text manually.

  • Ideal for structured input, where you already have a known delimiter.
  • Supports escape characters like \n (newline), \t (tab), and \r (carriage return).
  • More reliable than AI Mode since the split follows a precise rule.
  • If using GPT-generated input, you can instruct it to format the text with the required separator for easier splitting.

Separator Example

Example Use Cases​

  • Parsing CSV-like structured data.
  • Splitting user input fields formatted with a specific delimiter.
  • Handling predefined lists where items are separated by a consistent symbol.

Separator Example

Output Options: Lists & Loops​

Output as a List​

The output_as_list option allows the output to be wrapped inside a list.

  • When output_as_list is enabled, the node returns a single output containing a list of all split values.
  • Useful for nodes that expect grouped lists instead of separate items.

Using Data Splitter for Loops​

The Data Splitter Node integrates well with the Subflow Loop Node to process split data iteratively.

Example Workflow:​

  1. Enable output_as_list in the Data Splitter Node.
  2. Pass the resulting list into the Subflow Loop Node’s loop_over_fields input.
  3. The Subflow Loop processes each item sequentially, applying logic per element.

Loop Example

Best Practices​

  • Use AI Mode for short, clearly distinct inputs.
  • Use Separator Mode when dealing with structured text that includes known delimiters.
  • Enable output_as_list when downstream nodes require grouped lists (Subflow Loop).
  • Pair with the Subflow Loop Node to iterate over split elements dynamically.

Additional Resources​