JSON Converter Online

Convert JSON online

Pick a tab, then a target where there is a choice. All conversion runs in your browser; the page sends nothing to a server. There is no fixed size limit: the converters work on whatever your browser can hold, and very large inputs make the page slower.

• Convert → CSV — an array of objects becomes rows; keys become columns.
• Convert → YAML — the same data as YAML, with strings quoted where a YAML parser could read them as something else.
• Convert → XML — elements named after the keys, inside <root>; array members become repeated elements.
• Convert → SQL — one INSERT INTO data (…) VALUES (…); per object.
• Schema → TypeScript — interfaces inferred from the sample; keys missing from some objects become optional.
• Schema → JSON Schema — a draft 2020-12 schema with type, properties, required and items.
• Format and Minify — the same JSON, indented or on one line.

The single-purpose pages for two of these are the JSON to CSV converter and the JSON to TypeScript generator.

How to convert a JSON file

1. Click Upload and choose the .json file, drop it on the editor, or paste the text.
2. Pick the tab and the target format.
3. Click Copy, or Download to save it (data.csv, data.yaml, data.xml, data.sql, types.ts or schema.json).

The sample in the editor converts to this CSV:

id,name,role,tags
1,Ada,admin,"[""math"",""code""]"
2,Grace,editor,

Convert JSON to a readable format

Minified JSON is hard to read but means the same thing as indented JSON: RFC 8259 allows any amount of whitespace between tokens. The Format tab indents with 2 or 4 spaces and can sort keys A–Z; the JSON formatter does the same on its own page, and the minifier goes the other way.

How to open a .json file

A .json file is plain text, so any text editor opens it; RFC 8259 requires UTF-8 when JSON is exchanged between systems. To read it as a tree, use the JSON file viewer.

Convert invalid JSON to valid JSON

With Auto-repair input on (the default), broken input is repaired before it is converted, and the output pane is marked auto-repaired. These are the engine's results on common faults:

Two rows are guesses: NaN becomes null and 012 becomes the string "012". To see every change before converting, run the text through the JSON repair tool first, then check it in the validator. Parser messages by language are under JSON errors.

What JSON cannot convert to losslessly

• CSV is flat. A nested object or array is written into one cell as JSON text (the tags cell above), and the page says so under the output. A top-level array of numbers or strings cannot become CSV at all.
• CSV has no types. true, 1 and "1" all become plain text; null becomes an empty cell, the same as a missing key.
• Mixed keys. Columns are the union of all keys in first-seen order; a row without a key gets an empty cell.
• XML has no arrays or nulls. Array members become repeated elements and null becomes an empty element, so the way back to JSON is ambiguous.
• Large numbers. The input is parsed with JavaScript's JSON.parse, so integers beyond 2⁵³−1 lose precision before conversion; RFC 8259 notes that such numbers are not reliably exchanged.

Related JSON tools

XML to JSON converts the other way. JSON Schema generator, diff, the conversion guides (CSV, Python, YAML and TOML), Learn JSON and the full JSON Studio.

Frequently asked questions