JSON (JavaScript Object Notation) is a hierarchical, tree-structured data format. Objects can contain nested objects, arrays can hold other arrays, and a single JSON document can represent deeply layered data. This makes JSON ideal for APIs and configuration, but difficult to work with in spreadsheets and data analysis tools that expect flat rows and columns.

CSV (Comma-Separated Values) is a two-dimensional format: rows of data separated by newlines, with each row's fields separated by commas. There is no concept of nesting. Every tool from Microsoft Excel to Tableau to Python's pandas library can read CSV natively, making it the universal interchange format for tabular data.

JSON to CSV conversion is the process of flattening a JSON structure into a table. A simple JSON array of flat objects maps directly to rows and columns. Nested objects require a flattening strategy — the most common is dot notation, where { user: { name: 'Alice' } } becomes a column named user.name. Arrays within objects require an additional decision: join primitives as a single cell, or expand objects into multiple indexed columns.

The converter applies a consistent set of rules to turn any valid JSON array into a flat CSV table:

• Nested objects are expanded using dot notation. A field address.city in the object { address: { city: 'NYC', zip: '10001' } } produces two columns: address.city and address.zip.
• Arrays of primitives (strings, numbers) are joined into a single cell with a semicolon delimiter. The array tags: ['api', 'json', 'debug'] becomes the cell value api;json;debug.
• Arrays of objects are flattened with index notation. items[0].name, items[1].name, and so on become separate columns.
• Null values become empty CSV cells — not the string null.
• Missing fields across rows produce empty cells. Headers are collected from all records, so every possible column appears even if not every row has a value for it.

Values containing commas, double quotes, or newlines are automatically wrapped in double quotes per the RFC 4180 CSV standard. Double quotes within values are escaped by doubling them.

JSON to CSV conversion appears across a wide range of developer and data workflows. Understanding the common scenarios helps you choose the right flattening strategy for your data:

CSV is a simple format but has several subtleties that cause problems when opening files in spreadsheet tools. Understanding these helps you avoid common gotchas:

Browser-based JSON to CSV conversion covers the majority of real-world use cases, but some data structures require extra consideration:

• Non-uniform arrays. If different records have different sets of keys, the converter collects all unique keys and produces empty cells for missing fields. The resulting CSV is valid but may have many sparse columns.
• Deeply nested arrays of objects. Records with arrays-of-objects at multiple levels produce a large number of indexed columns (orders[0].items[0].sku). Consider pre-processing your JSON to flatten or select specific fields before converting.
• Top-level object instead of array. The converter expects a top-level JSON array. If your JSON is a single object, wrap it in array brackets: [{ ... }].
• Very large files. Converting files over 50 MB in the browser may be slow depending on your device. For bulk data processing, consider scripting with Node.js or Python's pandas library.
• Boolean and null values. Booleans become the strings true and false. Null values become empty cells. Both are correct CSV representations.