Must the JSON be an array?
An array of objects becomes one CSV row per object. A single object becomes one row.
Paste an array of objects and copy comma-separated rows for Excel, Sheets, or pandas. Conversion stays in this tab. Nested objects are stringified into cells unless you flatten them first or you want a real .xlsx from JSON to Excel.
JSON formatter
Run a tool to see a polished result here
Each object in an array becomes one CSV row. The union of keys becomes the header line. A single object is one row. Primitive arrays become one column of values if you wrap them as objects; a raw array of numbers is not a table. Nested objects and nested arrays are JSON-stringified into a single cell โ that is the important difference from JSON to Excelโs flatten mode.
Need a real workbook with dotted columns like address.city? JSON to Excel flattens nested objects and downloads .xlsx.
Need elements rather than a grid? JSON to XML is the converter that maps objects to XML.
Garbage CSV usually means the input was not JSON. Check it with the JSON validator before converting.
A value like `"tags": ["json","csv"]` becomes one CSV cell containing the JSON array, quoted so the inner commas do not create extra columns. A value like `"address": {"city":"London"}` becomes one cell containing `{"city":"London"}`. If you need tags.0 / tags.1 or address.city as headers, switch to JSON to Excel with flatten, or pre-process the JSON so those fields are already top-level.
[ { "name": "Ada", "tags": ["json", "csv"] } ]
โ name,tags
Ada,"[""json"",""csv""]"Row 1 has `role`, row 2 does not: the header still includes role, row 2 leaves it empty. `null` becomes an empty cell, not the text null. Booleans become `true`/`false` text. Numbers stay unquoted. Strings that contain commas, quotes, or line breaks are RFC 4180-quoted.
Excel may split UTF-8 CSV on the wrong delimiter if a BOM is missing; this converter emits UTF-8 without a BOM. Cells that look like `=SUM(A1)` are still CSV text โ Excel may treat them as formulas when opened. For typed cells and no formula injection, download .xlsx from JSON to Excel instead. Very wide objects create very wide CSV; that is still one header line.
The sample is two people with nested address objects and tag arrays. CSV mode will put `{"city":"London"}` into an address cell and `["json","csv"]` into a tags cell. That is expected: this converter does not explode nested keys. Compare with the same sample on JSON to Excel using flatten.
[
{ "name": "Ada", "role": "engineer", "address": { "city": "London" }, "tags": ["json", "csv"] },
{ "name": "Linus", "role": "maintainer", "address": { "city": "Helsinki" }, "tags": ["kernel"] }
]Use JSON to CSV when the output must be portable text: git diffs, pandas.read_csv, database imports. Use JSON to Excel when someone needs to double-click a .xlsx. Do not use this page to convert XML or YAML โ those have their own URLs.
Also try JSON to Excel, JSON to XML, or JSON โ YAML. The JSON formatter on the homepage is the general workspace.
An array of objects becomes one CSV row per object. A single object becomes one row.
Not on this page. Nested values are stringified into cells. Use JSON to Excel flatten mode for dotted columns.
Usually. There is no BOM. If glyphs break, import the file as UTF-8 or use the .xlsx downloader instead.
Fields that contain commas, quotes, or newlines are wrapped in double quotes; inner quotes are doubled.
No. Conversion stays in this tab.