Loop

Repeat actions for each item in a list

What does this node do?

The Loop node processes arrays by repeating a sequence of nodes for each item. It’s essential for batch operations like scraping multiple URLs, processing spreadsheet rows, or handling API response lists.

Common uses:

• Process a list of URLs
• Iterate through spreadsheet rows
• Handle API response arrays
• Batch process data items

Quick setup

Configuration

Required fields

Optional fields

Loop variables

Inside a loop, access these special variables:

Variable	Description	Example
{{Loop_0.currentItem}}	Current item	"https://example.com"
{{Loop_0.index}}	Current index (0-based)	0, 1, 2…
{{Loop_0.totalItems}}	Total item count	10

Accessing item properties

If items are objects:

// Items array
[
  {"url": "https://a.com", "name": "Site A"},
  {"url": "https://b.com", "name": "Site B"}
]

Access properties:

{{Loop_0.currentItem.url}}   → "https://a.com"
{{Loop_0.currentItem.name}}  → "Site A"

Output

After completion, the loop provides:

{
  "results": [
    // Array of results from each iteration
  ],
  "totalItems": 10,
  "processedItems": 10,
  "errors": []
}

Access results: {{Loop_0.results}}

Examples

Scrape multiple URLs

Input: List of URLs from Google Sheets

graph LR
    A[Google Sheets] --> B[Loop]
    B --> C[Web Scraper]
    C --> D[LLM Analysis]
    D --> E[Save Results]

Configuration:

• Items: {{GoogleSheets_0.data}}
• Max iterations: 50
• Delay: 2000ms

Inside loop:

• Web Scraper URL: {{Loop_0.currentItem.url}}

Process API results

Input: API response with array of records

graph LR
    A[HTTP Request] --> B[Loop]
    B --> C[Transform Data]
    C --> D[BigQuery Write]

Configuration:

• Items: {{HTTP_0.body.results}}

Email multiple recipients

Input: Contact list

graph LR
    A[Contacts] --> B[Loop]
    B --> C[LLM Personalize]
    C --> D[Email Sender]
    D --> E[Log Sent]

Inside loop:

• Email to: {{Loop_0.currentItem.email}}
• Name variable: {{Loop_0.currentItem.name}}

Nested loops

Process multi-dimensional data:

graph LR
    A[Categories] --> B[Outer Loop]
    B --> C[Products per Category]
    C --> D[Inner Loop]
    D --> E[Process Product]

Configuration:

• Outer loop: {{categories}}
• Inner loop: {{OuterLoop_0.currentItem.products}}

Best practices

Always set max iterations

Prevent infinite or runaway loops:

Max iterations: 100

Even if you expect fewer items, set a limit.

Add delays for external calls

When calling APIs or scraping:

Delay: 2000 ms (2 seconds)

This prevents rate limiting and server overload.

Handle errors in the loop

Add error handling for each iteration:

graph LR
    A[Loop Item] --> B{Try Operation}
    B -->|Success| C[Save Result]
    B -->|Error| D[Log Error]
    C --> E[Next Item]
    D --> E

Use Conditional nodes to check for failures.

Batch large datasets

For thousands of items:

1. Split into batches of 50-100
2. Process each batch in a loop
3. Add delays between batches
4. Track progress in a log

Use meaningful node names

Inside loops, rename nodes clearly:

❌ WebScraper_0
✅ ScrapeCurrentURL

❌ LLM_0
✅ AnalyzeContent

Common patterns

Collect and aggregate

Loop through items
  → Process each
  → Add to results array

After loop:
  → Aggregate results
  → Generate summary

Filter while processing

Loop through items
  → Check condition
  → If matches, process
  → If not, skip

Progress tracking

Loop through items
  → Process
  → Log progress: "Processed {{Loop_0.index + 1}} of {{Loop_0.totalItems}}"

Common issues

Loop runs too long

• Reduce max iterations
• Add shorter delays
• Process in smaller batches
• Check for infinite loop conditions

Rate limited by API

• Increase delay between iterations
• Reduce batch size
• Implement exponential backoff
• Check API rate limits

Some items fail

• Add Conditional node to handle errors
• Log failed items for review
• Use retry logic for transient failures

Results array is empty

• Ensure nodes inside loop produce output
• Check that results are being collected
• Verify input array has items

Related nodes