CSV to JSON Converter

CSV to JSON Converter: Turn a Spreadsheet Export Into Real, Structured Data

CSV is a great way to move data out of a spreadsheet, but it's a poor way to move data into a program. Every value in a CSV is just text separated by commas, with no sense of numbers, booleans, or nested structure, and no standard way to represent a field that's genuinely empty. JSON solves all of that: it has real types, a predictable shape, and it's the format every modern web app, REST API, and NoSQL database already speaks natively. Converting once, correctly, saves rewriting the same parsing logic in every project that needs the data.

Every converter on SmallStudyTools is tested with real files before it ships. This tool auto-detects your CSV's delimiter (comma, semicolon, or tab), parses it with a proper character-by-character reader that correctly handles quoted fields with embedded commas and newlines, and builds valid JSON from the result, no shortcuts and no silently mangled rows.

1How This CSV to JSON Converter Works

The tool reads your file as text, detects which delimiter it actually uses by sampling the first line, and parses every row with a small state machine that tracks whether it's currently inside a quoted field, so a comma or line break sitting inside quotes never breaks the row apart. The first row becomes your column headers, and every row after it becomes a JSON value shaped by whichever Output Structure you've chosen. The result is validated JSON, ready to paste straight into code, an API tester, or a database import tool.

2Simple vs Pro Mode

Simple Mode covers the everyday case: upload one CSV and get back a clean array of JSON objects, 2-space indented, with numbers and true/false values already detected, ready to use immediately. Pro Mode, currently free to use, hands over full control of the output shape and formatting, plus bulk conversion for anyone processing a whole folder of exports at once.

3Output Structure, Explained

This is the option worth understanding before converting anything, since it decides what the JSON actually looks like. Array of Objects, the default, gives you [{ "name": "Alice", "age": 30 }, ...], one object per row with the header row as keys, which is what most web apps, form libraries, and APIs expect. Array of Arrays keeps things closer to the original spreadsheet shape, header included as its own first row, useful when a script needs to process the data positionally rather than by key. Keyed by First Column turns the first column's values into top-level object keys instead of array indexes, which is a natural fit for a CSV where that column is a unique ID, a SKU, or a username, letting you look a record up directly by that value instead of searching an array. NDJSON is the odd one out: instead of one JSON value for the whole file, it writes one compact object per line with nothing wrapping them, covered in more detail below.

4NDJSON: One Object Per Line

NDJSON, short for newline-delimited JSON and also known as JSON Lines, writes each row as its own complete, self-contained JSON object on its own line, with no enclosing square brackets and no commas joining one record to the next. That shape matters for tools that process data as a stream rather than loading a whole file into memory at once: log processors, many data pipeline tools, and the bulk import endpoints of systems like Elasticsearch and BigQuery are all built to read one line, act on it, and move to the next, which a single giant JSON array doesn't support cleanly. If your export is headed for one of those systems, or you're feeding records into a pipeline one at a time, NDJSON is usually the format it's actually asking for even when the request just says "JSON." Because each line stands alone, the Indentation setting doesn't apply to NDJSON output, every line stays compact, and downloaded files use the real .ndjson extension rather than .json.

5Common Reasons People Convert CSV to JSON

Most CSV to JSON conversions happen at the boundary between "data someone exported from a spreadsheet" and "data a program needs to consume." A marketing list exported from a CRM that needs to become the seed data for a web app, a product catalog headed into a NoSQL database that only speaks JSON documents, or a batch of records being tested against an API endpoint before writing any integration code are all common starting points. Since JSON is a first-class data type in JavaScript and nearly every modern backend language, converting once removes an entire category of manual parsing work.

6Doing It in JavaScript vs Using This Tool

It's genuinely common to reach for a quick script instead of a tool, and for a very small, simple CSV, a naive JavaScript conversion can look deceptively easy:

// A naive version — breaks on quoted fields containing commas
const rows = csvText.trim().split('\n').map(r => r.split(','));
const headers = rows[0];
const json = rows.slice(1).map(r =>
  Object.fromEntries(headers.map((h, i) => [h, r[i]]))
);

That snippet works right up until a cell contains a comma inside quotes, a value spans multiple lines, or a number needs to come through as an actual number instead of text, at which point the "quick" script needs a real parser, type coercion, and edge-case handling to match what this tool already does. Pasting the result into the Output Structure and Key Naming options here gets you the same outcome as a hand-written script, minus the debugging, and the JSON it produces is safe to drop straight into any JavaScript project, fetch mock, or test fixture.

7Type Inference and Empty Values

CSV has no concept of data types, every cell is just text, so "30" and "true" arrive looking identical to "Alice". Type Inference fixes that by converting cells that look like whole numbers, decimals, or the words true/false into real JSON numbers and booleans. Empty Values is a separate, independent setting for the specific case of a blank cell: keep it as an empty string exactly as the CSV had it, convert it to JSON's explicit null, or omit the field from the object entirely, which keeps the JSON smaller and is often what a schema with optional fields actually wants instead of a string of empty placeholders.

8Key Naming: camelCase Headers

Spreadsheet headers like "First Name" or "Customer ID" are readable, but they're awkward as JavaScript object keys, since they need bracket notation and don't match the naming convention most JS codebases use. Turning on Key Naming rewrites every header into camelCase automatically, so "First Name" becomes firstName and "user_id" becomes userId, letting the converted JSON drop straight into a codebase using standard dot notation without a manual rename pass.

9Frequently Asked Questions