CSV to JSON Converter — Free Online Tool

When you'll convert CSV to JSON

CSV is the workhorse of business data exports — spreadsheets, CRM dumps, BI exports, analytics downloads. But almost every modern API, JavaScript framework and NoSQL database wants JSON. Converting CSV to JSON is the bridge that lets analytics-team output flow into the engineering side of the stack.

Common scenarios: loading a CSV of customers into a Node.js script for processing; importing a sales export into MongoDB or Firestore; feeding survey results into a React app for visualisation; turning a spreadsheet of products into a mock JSON API for prototyping; converting time-series exports for chart libraries that expect JSON.

How this converter works

The tool parses your CSV using a proper RFC 4180 parser that handles all the edge cases: quoted fields with embedded commas, quoted fields with embedded newlines, escaped double-quotes ("" inside a quoted string). The result is an array of objects where each row becomes one object, and the keys are taken from the header row.

If "Infer types" is on, the parser tries to convert values: 42 becomes a number, true/false become booleans, null becomes null. Everything else stays as a string. Turn this off if you need every field as a string (useful when "01234" should stay as a string and not become the number 1234).

Delimiter auto-detection

"Auto-detect" inspects the first line and counts commas, semicolons, tabs and pipes — picking the most-frequent one as the delimiter. Works correctly in 95% of cases. If your CSV has unusual content (e.g. text fields containing many commas), pick the delimiter manually.

Type inference rules

• Numbers — strings matching /^-?\d+(\.\d+)?$/ become numbers.
• Booleans — exact strings true or false (case-insensitive) become booleans.
• Null — exact string null or empty cell becomes null.
• Everything else stays a string.

Common gotchas: leading-zero IDs (01234) become numbers (1234) with inference on — turn it off to preserve the leading zero. Same for phone numbers, postcodes and ZIP codes.

Without a header row

Untick "First row is header" and column keys become field1, field2, etc. Use this for CSVs that don't have headers.

Tips and best practice

• Verify the auto-detected delimiter on the first parse. If columns look wrong, switch it manually.
• Watch leading zeros and big numbers — JSON numbers don't preserve leading zeros and can lose precision beyond 2^53. Turn off type inference for IDs.
• For huge CSVs (100k+ rows), browser parsing may take several seconds. Consider a streaming parser in Node.js for production pipelines.
• If a value contains both quotes and commas, your source CSV must quote it correctly: "a, ""b"", c". The parser handles this; bad source files will produce odd results.

FAQs

Related CSV tools