What JSON formats can I convert to?

The CSV / YAML tab exports JSON in two directions: JSON → CSV — for arrays of flat objects (string, number and boolean values only). The first row is the header, fields are quoted and escaped correctly, and a UTF-8 BOM is prepended so Excel opens it cleanly. Nested objects or arrays are rejected with a clear message. JSON → YAML — any JSON structure, produced locally with the yaml library and downloadable as .yaml .

What JSONPath syntax is supported?

The JSONPath tab supports the commonly used subset: the root $ , dot child access ( $.users.name ), bracket access with quoted keys ( $['users']['name'] ) and numeric indexes ( $[0] ), wildcards ( $.users[*] ), union selection ( $[0,2] ), array slices ( $[1:3] , $[:2] ), and recursive descent ( $..id ). Run a query to see every matching value with its index, or copy the full list of matches as JSON.

JSON Formatter & Validator Online

Format, validate, convert, diff, infer schemas, and query JSON — all locally in your browser.

Result

Tree view renders a collapsible hierarchy of the parsed JSON so you can explore nested arrays and objects without reading braces. Format first, then switch to Tree.

Input

What is JSON Formatter & Validator?

A browser-based workbench for JSON developers and QA engineers. Every operation runs client-side with the native JSON.parse, so pasted data never leaves your device.

Format & Validate — pretty-print with 2 spaces or tabs, minify, and run a dedicated Validate check with a detailed error panel (message, offset, line, column).

Convert — export JSON to CSV (arrays of flat objects) or YAML, with one-click copy and download.

Diff — compare two JSON documents at token level; unchanged/added/removed tokens are color-coded with their JSONPath location.

Schema Infer — derive a JSON Schema draft 2020-12 document from sample data, including anyOf for mixed types and required for always-present keys.

JSONPath — query documents with $, dot/bracket children, wildcards, slices, and recursive descent, then copy the matching values.

Zero-Server Tool Data Guarantee

Everything runs locally in your browser using JavaScript's native JSON.parse. No JSON data is ever sent to a server, logged, or cached — the yaml, CSV, diff, schema, and JSONPath conversions all happen on your device.

How to Use

Pick a tab

Choose from Format & Validate, CSV / YAML, Diff, Schema Infer, or JSONPath.

Paste your JSON

Type or paste any valid (or not yet valid!) JSON into the input area.

Run the operation

Format, Validate, Convert, Compare, Infer Schema, or Run Query — results appear instantly.

Copy or download

Grab formatted output, schemas, matches, CSV, or YAML via the copy button or download link.

Common Use Cases

Debugging API Responses

Paste a failing API response, format it, validate it, and jump straight to the error position instead of squinting at one long line.

Contract Testing

Compare two versions of a payload with the token-level diff to confirm exactly which fields changed before updating mocks or fixtures.

Schema-First Development

Feed sample records into Schema Infer to bootstrap a JSON Schema draft 2020-12 document, then tighten constraints by hand.

Data Migration Prep

Export a flat array of objects to CSV for spreadsheet review, or to YAML for configuration and CI workflows.

Log / Telemetry Mining

Use JSONPath to pull specific fields out of nested log lines without writing a full parser.

Implementation Examples

JavaScriptValidate + JSONPath

const raw = '{"users":[{"id":1,"name":"Ada"}]}';
let data;
try {
    data = JSON.parse(raw);
    console.log("Valid JSON");
} catch (e) {
    console.error(`Invalid: ${e.message}`);
}

// JSONPath-style query (manual)
if (data) {
    const names = data.users.map(u => u.name);
    console.log(names); // ["Ada"]
}

JSON Tooling Approaches

Approach	Where it runs	Best for	Notes
This browser toolbox	Client-side	Quick interactive checks	Free, private, no setup — yaml/CSV/schema/diff/JSONPath included
jq (CLI)	Local machine	Scripted filtering of files	Powerful filters, streaming, but no GUI
jq + yq combo	Local machine	JSON + YAML scripting	Two tools to learn; yq mirrors jq syntax
Language libraries (json, yaml)	In your app	Runtime parsing in code	Not interactive; needs a project setup
Online-only validators	3rd-party server	One-off checks	Your data leaves the browser; often ad-funded

Production Best Practices & Security

Validate before you format — use the Validate button to surface the exact error position (line and column) before pasting data into production configs. Why:a single misplaced comma or unescaped quote is the most common cause of JSON.parse failures in automated pipelines.

Use CSV only for flat object arrays — the CSV export intentionally rejects nested objects and arrays. Why:CSV has no native representation for nested structures, and flattening them silently corrupts the data.

Compare with the same key order — the token-level diff treats a reordered key as removed + added. Why:object key order is semantically irrelevant in JSON, so sort keys consistently before diffing to see only real changes.

Treat inferred schemas as a starting point — inferred schemas capture observed structure, not intent. Why:a schema that permits only what you saw will reject valid future data; always add constraints like enum, pattern, and range afterwards.

Keep queries scoped — start JSONPath expressions narrow (e.g. $.users[0]) and widen only when needed. Why:recursive descent (..) walks the entire tree and can return far more matches than expected.

Never embed secrets — avoid pasting private keys or tokens into any online tool. Why:even client-side tools can be compromised; use local CLIs for real key material.

Frequently Asked Questions

It bundles five JSON development checks into one page: Format & Validate (pretty-print with 2 spaces or tabs, minify, and a Validate button with a detailed error panel), CSV / YAML conversion, token-level diff between two documents, JSON Schema inference (draft 2020-12), and JSONPath queries. Everything runs in your browser using the native JSON.parse parser.

Yes. Clicking Validate runs the input through JavaScript's native JSON.parse — the same strict RFC 8259 parser used everywhere. Valid input gets a green confirmation with a quick summary (node count, maximum depth, and size in KB).

Invalid input opens a red error panel showing the parser's message, the byte offset, and the exact line and column where parsing stopped, so you can jump straight to the problem instead of scanning the whole document.

The Tree view renders your parsed JSON as a collapsible hierarchy instead of raw braces and brackets. Arrays and objects can be expanded or collapsed, keys are shown in document order, and values are color-coded by type (strings, numbers, booleans, null).

Format your JSON first, then switch to the Tree view to explore deeply nested API responses or config files without reading a single line of punctuation.

There is no hard-coded size limit — the tool uses your browser's native JSON parser which can handle large payloads.

However, browser memory constraints apply (typically 1-4 GB depending on your system). JSON payloads up to 10 MB process instantly. Files in the 50-100 MB range may cause noticeable slowdowns.

For production-scale data transformation (gigabyte-sized files), use Node.js or a dedicated CLI tool like jq.

The CSV / YAML tab exports JSON in two directions:

JSON → CSV — for arrays of flat objects (string, number and boolean values only). The first row is the header, fields are quoted and escaped correctly, and a UTF-8 BOM is prepended so Excel opens it cleanly. Nested objects or arrays are rejected with a clear message.
JSON → YAML — any JSON structure, produced locally with the yaml library and downloadable as .yaml.

The Diff tab compares two JSON documents token by token — keys, string values, numbers, booleans, null, and punctuation. Tokens that appear in both documents are left unchanged, removed tokens are struck through in red, and new tokens are highlighted in amber, each annotated with its JSONPath location via the tooltip.

Because the comparison runs over tokens, a reordered property is reported as removed + added rather than a move. For very large documents the comparison is capped so your browser stays responsive.

Yes, on both counts. Formatting, validation, CSV/YAML conversion, diffing, schema inference, and JSONPath queries all run client-side — your data never leaves your device, is never transmitted, and is never cached on a server.

For the token-level diff, only your browser computes the comparison; for schema inference, only your browser derives the schema. The tool is free with no account or API key required.