Paste or upload a CSV file and get clean, formatted JSON instantly — live preview, no upload required.
CSV (Comma-Separated Values) and JSON (JavaScript Object Notation) are the two most widely used data interchange formats on the web. CSV dominates the world of spreadsheets, databases, and data exports — it is human-readable, compact, and universally supported by tools like Microsoft Excel, Google Sheets, PostgreSQL, and virtually every data pipeline. JSON, on the other hand, is the lingua franca of modern APIs, NoSQL databases, and JavaScript applications. When you need to move data from one world to the other, a reliable CSV to JSON converter is indispensable. Conversion sounds simple on the surface — take rows and columns, turn them into objects and arrays — but the details matter enormously in practice. Real-world CSV files are rarely clean: they use semicolons instead of commas (common in European locales), values contain embedded commas or newlines inside quotes, cells hold numeric-looking strings that should stay as strings or be cast to numbers, boolean flags appear as the literal text 'true' or 'false', and column headers follow a 'address/city' slash notation that implies a nested JSON object structure. A robust converter must handle all of these edge cases gracefully, and that is exactly what this tool is built to do. This CSV to JSON Converter uses PapaParse — the same industry-standard parsing library relied on by thousands of open-source projects including jam.dev's utilities — to handle the full RFC 4180 CSV specification. PapaParse correctly deals with multi-line quoted fields, byte-order marks (BOM) from Excel exports, and delimiter ambiguity. On top of the parser, this tool layers a rich set of conversion options that let you control every aspect of the output JSON. Common use cases include: preparing data for MongoDB or Elasticsearch import by converting CSVs to NDJSON (newline-delimited JSON, one record per line); transforming product catalog exports from an e-commerce platform into JSON arrays for API ingestion; converting survey responses from Google Forms CSV exports into JSON objects for analysis in Node.js or Python; building configuration files from spreadsheet-managed data; and seeding databases with test fixtures exported from relational database dumps. The live preview feature means you see the JSON output update in real time — within 300 milliseconds of any change to the input or any option toggle. You do not need to click a 'Convert' button. This dramatically speeds up the workflow when you are iterating on delimiter settings or checking whether parse-numbers is doing what you expect. Privacy is built in by design. This tool is 100% browser-based. Your CSV data is processed entirely in your local JavaScript runtime and never transmitted to any server. This makes it safe to use with sensitive datasets — financial records, employee data, customer lists, health records — without any risk of data exposure. The output can be formatted as a pretty-printed JSON array of objects (the most common format, with configurable 2-space, 4-space, or tab indentation), as a minified single-line JSON string for embedding in code, as a keyed hash map where the first column value becomes the top-level key, or as NDJSON (JSON Lines) for streaming and database import pipelines. You can also toggle lowercase key normalization, nested object creation via slash notation in headers, and empty cell handling (empty string vs JSON null).
Understanding CSV to JSON Conversion
What Is CSV to JSON Conversion?
CSV to JSON conversion is the process of transforming tabular data — rows and columns separated by a delimiter character — into hierarchical JSON objects. In the standard 'array of objects' mode, each CSV row becomes a JSON object where the keys are taken from the first (header) row and the values come from the corresponding data cells. The entire dataset becomes a top-level JSON array. For example, a three-column, two-row CSV with headers 'name', 'age', and 'city' becomes a JSON array containing two objects, each with those three keys. This format is directly consumable by JavaScript code, REST APIs, NoSQL databases, and most modern data processing tools.
How Does the Conversion Work?
The converter uses PapaParse to tokenize the CSV input according to RFC 4180 rules, correctly handling quoted fields that contain commas or newlines, doubled double-quotes as escape sequences, and optional BOM characters at the start of Excel-exported files. Once tokenized into a two-dimensional array of strings, the tool applies your chosen options: the first row is used as header keys (or auto-generated numeric field names if disabled), values are optionally trimmed of whitespace, numeric strings are optionally cast to JavaScript numbers, boolean literals are optionally cast to true/false, empty cells become empty strings or JSON null, and slash-notated header keys like 'address/city' are split into nested object paths. The resulting object array is then serialized with JSON.stringify using your chosen indentation setting.
Why Does This Conversion Matter?
JSON is the standard input format for modern APIs, NoSQL databases like MongoDB and Firestore, and JavaScript/TypeScript applications. CSV is the standard export format for spreadsheets, relational databases, and legacy enterprise systems. The gap between them is wide and frequently encountered: a developer receives a product catalog as a CSV from a client and needs it as a JSON array to seed a database; a data analyst exports query results as CSV and needs NDJSON to stream into Elasticsearch; a devops engineer manages configuration in a shared spreadsheet and needs to convert it to JSON config files. Doing this manually is error-prone and slow — a reliable, browser-based converter eliminates the friction entirely.
Limitations and Edge Cases
This tool processes CSV files up to 5 MB in the browser — larger files may cause performance issues and are better handled by server-side tools or command-line utilities like csvkit or jq. The nested object feature (slash notation) handles one level of nesting but does not support array syntax or multiple nesting levels beyond path splitting on '/'. The parse-numbers option converts any value that passes JavaScript's Number() coercion, which means values like '007' become 7 (losing the leading zero) — if you need to preserve leading zeros, keep parse-numbers disabled. The keyed output mode uses the first column as the key, so duplicate first-column values will overwrite earlier records. NDJSON output is always minified (one compact object per line) regardless of the indentation setting.
How to Convert CSV to JSON
Paste or Upload Your CSV
Click the 'Paste CSV' tab and paste your CSV data directly into the text area, or switch to 'Upload File' to browse and select a .csv, .tsv, or .txt file from your device. You can also click 'Load Sample' to try the tool with example data before using your own file.
Configure Delimiter and Options
Select your delimiter — leave it on 'Auto' to have the tool detect commas, semicolons, tabs, or pipes automatically. Toggle 'First row as keys' if your CSV has a header row (recommended). Enable 'Parse numbers' to convert numeric strings to JSON numbers, or 'Parse booleans & null' to cast true/false/null literals to their JSON equivalents.
Review the Live JSON Output
The JSON output updates automatically as you type or change any option — no need to click a Convert button. Switch between 'Array of Objects' (default), 'Keyed Object' (hash map using first column as key), or 'NDJSON / JSON Lines' (one record per line, ideal for MongoDB imports). Click 'Show table' to see a preview of your parsed CSV rows before the JSON.
Copy or Download the Result
Click 'Copy JSON' to copy the full JSON output to your clipboard, or click 'Download .json' to save it as a file. The filename is derived from your uploaded file name when available. The conversion summary shows the record count, column count, and estimated JSON file size.
Frequently Asked Questions
What delimiters does the converter support?
The converter supports comma (,), semicolon (;), tab character, pipe (|), and colon (:) as explicit delimiter choices. The 'Auto' setting uses PapaParse's built-in delimiter detection, which counts occurrences of common delimiter characters in the first line and picks the one that appears most frequently. Auto-detection works well for the vast majority of CSV and TSV files exported by spreadsheet applications and databases. If your file uses a more unusual delimiter (such as a caret ^ or tilde ~), you can open the file in a text editor, replace the delimiter with a supported one, and then paste the modified content into the tool.
How do nested objects work with the slash notation?
When the 'Nested objects (/)' option is enabled, any column header containing a forward slash is interpreted as a nested path. For example, a header named 'address/city' produces { "address": { "city": "..." } } in the output object. Multiple headers sharing the same prefix — such as 'address/city' and 'address/zip' — are merged into the same nested object: { "address": { "city": "...", "zip": "..." } }. This is useful when you receive flat CSV exports from systems that encode hierarchical data as dot-or-slash-notated column names. The feature only handles one level of nesting per slash segment — deeply nested paths like 'a/b/c' produce { "a": { "b": { "c": "..." } } }.
Can I convert large CSV files?
This tool supports files up to 5 MB, which covers most practical use cases — a 5 MB CSV can contain tens of thousands of rows depending on column count and value lengths. Because all processing happens in your browser's JavaScript runtime, very large files (10 MB or more) may cause noticeable slowness or run out of memory on low-end devices. For files larger than 5 MB, consider splitting the file into chunks using a tool like csvsplit, or use a command-line converter such as csvkit's csvjson command or a Node.js script using the papaparse npm package directly, which supports streaming large files via web workers.
Is my data kept private? Does it get sent to a server?
Your data never leaves your device. This CSV to JSON converter is entirely browser-based — all parsing and conversion is performed by JavaScript running locally in your browser tab. No CSV content, no converted JSON, and no metadata is transmitted to any server. This makes the tool safe to use with sensitive or confidential data including financial records, employee information, customer lists, health data, and proprietary business data. You can verify this by checking your browser's network activity tab (F12 → Network) and observing that no outbound requests are made when you paste or convert data.
What is NDJSON mode and when should I use it?
NDJSON stands for Newline-Delimited JSON, also called JSON Lines. In this format, each JSON object is written on its own line with no surrounding array brackets and no commas between records. For example: {"name":"Alice"} {"name":"Bob"}. This format is the standard import format for MongoDB's mongoimport tool, Elasticsearch's bulk API, and many streaming data pipeline systems like Kafka and Logstash. It is also useful for large datasets because each line can be parsed independently without loading the entire file into memory. Use NDJSON mode when your target system expects one JSON object per line rather than a JSON array.
Why do some values stay as strings even with Parse Numbers enabled?
The parse-numbers option converts a value to a JSON number only if the entire value passes JavaScript's Number() coercion and the result is not NaN. Values that look like phone numbers (e.g., '0123456789'), zip codes ('01234'), product codes ('007'), or version strings ('1.2.3') may be converted to numbers, losing leading zeros or being rejected entirely. If you need to preserve leading zeros or keep numeric-looking strings as strings, leave parse-numbers disabled. The option is best used for columns that contain genuine numeric data such as prices, quantities, ages, or scores where leading-zero preservation is not needed.