JSON to CSV in Power Automate, n8n, and Zapier
- Power Automate: use the "Select" action to map JSON fields, then "Create CSV table"
- n8n: use the "Spreadsheet File" node or "Convert to CSV" operation
- Zapier: use the Formatter > Utilities > Line Itemizer step
- For one-off conversions, the browser tool is faster than any no-code setup
Table of Contents
For automated workflows that repeatedly convert JSON to CSV as part of a pipeline, Power Automate, n8n, and Zapier all have native ways to handle this. For a one-time or occasional conversion, the browser-based JSON to CSV converter is faster to set up than any automation platform — paste, convert, download, done.
Here's how each no-code platform approaches JSON to CSV, and when each is the right choice.
Power Automate: Create CSV Table
Power Automate has a built-in "Create CSV table" action that converts a JSON array to CSV format inside a flow.
Basic setup:
- Parse your JSON using the "Parse JSON" action (provide a schema based on your data structure)
- Add a "Create CSV table" action — set "From" to the array output from Parse JSON
- Choose "Automatic" for columns (Power Automate infers headers from the first object) or "Custom" to map specific fields
- The action outputs a CSV string — save it to SharePoint, OneDrive, or send via email
Nested object limitation: Power Automate's "Create CSV table" doesn't flatten nested objects automatically. If your JSON has nested fields like address.city, you'll need a "Select" action before the CSV step to explicitly map each field to a flat key name.
When to use: Recurring workflows where JSON arrives from a trigger (HTTP request, SharePoint update, Dataverse record) and needs to become a CSV on a schedule.
n8n: Convert JSON to CSV
n8n offers multiple paths to convert JSON data to CSV:
Method 1: Spreadsheet File node (recommended)
- Add a "Spreadsheet File" node after your data-returning node
- Set Operation to "Write" and File Format to "csv"
- The node converts n8n's item array to CSV automatically
- Connect to a "Write Binary File" node to save, or pass to email/storage nodes
Method 2: Code node for custom handling
For complex transformations (custom headers, nested flattening), add a Code node and write the CSV string manually using JavaScript array manipulation before passing to a Write Binary File node.
When to use: n8n is ideal for self-hosted automation where you need full control over data transformation. The visual workflow builder makes it easy to see where JSON is being converted in your pipeline.
Sell Custom Apparel — We Handle Printing & Free ShippingZapier: JSON Array to CSV via Formatter
Zapier doesn't have a direct "JSON to CSV" action, but the Formatter app with Utilities can handle simple cases:
- Add a "Formatter by Zapier" step
- Choose "Utilities" > "Line Itemizer" to split a JSON array into individual lines
- Use subsequent Formatter steps to structure the output as CSV rows
This works for flat JSON arrays but gets complex quickly with nested data. For anything beyond simple key-value pairs, Zapier's approach becomes unwieldy — consider n8n or Power Automate for JSON-heavy workflows.
Zapier also integrates with Google Sheets directly: if your end goal is to add JSON data to a Sheets spreadsheet, use the "Google Sheets: Create Spreadsheet Row" action and map fields individually — avoiding the CSV intermediate step entirely.
Make.com: Create CSV Module
Make.com (formerly Integromat) has a "Create CSV" module under the CSV app. It takes an array of objects and produces a CSV file:
- Add the "CSV > Create CSV" module
- Map each field from your JSON to a CSV column — you define headers and which JSON values go in each column
- The output is a CSV file that can be saved, emailed, or passed to another module
Make.com's CSV module is more flexible than Zapier's Formatter but requires more setup than n8n. For teams already using Make.com, it's the right native option.
When to Skip the Automation and Just Use the Browser
All four platforms (Power Automate, n8n, Zapier, Make.com) require setup time: connecting accounts, mapping fields, testing triggers, handling errors. For recurring automated workflows, that investment pays off. For one-off or occasional conversions, it's pure overhead.
Use the browser tool instead when:
- You have a JSON file from an API response, database export, or coworker — and you just need CSV right now
- You're testing a data structure before committing to an automated pipeline
- Your automation platform can't handle the nested structure cleanly and you need to pre-flatten
- You're not technical enough to configure a Power Automate flow but need to get the job done
Paste JSON, click Convert, download CSV. No triggers, no mapping steps, no accounts. For everything else, set up the automation.
Not Building an Automation? Convert JSON to CSV Right Now
No triggers to configure, no field mapping, no account. Paste your JSON array, download CSV. Takes 30 seconds.
Open Free JSON to CSV ConverterFrequently Asked Questions
Can Power Automate handle nested JSON when creating a CSV table?
Not automatically. The "Create CSV table" action flattens one level of object properties. For nested objects, add a "Select" action first to explicitly map each nested field to a flat key name, then pipe that into "Create CSV table."
What is the best no-code tool for JSON to CSV conversion in a workflow?
n8n has the most direct JSON-to-CSV path with the Spreadsheet File node. Power Automate works well for Microsoft 365 environments. Zapier is weakest for structured data transformation — consider it last if JSON-to-CSV is central to your workflow.
Does Make.com have a free plan that includes CSV conversion?
Yes, Make.com's free plan includes 1,000 operations per month. The CSV module is available on the free tier. Each CSV creation counts as one operation.
Can I test my JSON structure before building the automation?
Yes — paste your sample JSON into the browser tool first to verify the CSV output looks correct. This lets you confirm column headers, nesting behavior, and data quality before investing time in the automation setup.

