How to create a CSV file.

A CSV is the simplest possible way to represent a table of data: each row on its own line, each column separated by a comma. The first line is usually headers. Every spreadsheet and every programming language reads it. Four ways to make one — pick the one that matches where your data already lives.

1. From a spreadsheet (easiest)

Excel, Google Sheets, Apple Numbers, and LibreOffice Calc all export to CSV in two clicks:

• Excel: File → Save As → choose "CSV UTF-8 (Comma delimited) (*.csv)". Plain "CSV (Comma delimited)" uses Windows-1252 and will mangle non-ASCII. Always pick the UTF-8 variant.
• Google Sheets: File → Download → Comma-separated values (.csv).
• Apple Numbers: File → Export To → CSV.

Use this when the data is already in a spreadsheet. No escaping or formatting to get wrong — the app handles all of it.

2. From a text editor (for tiny files)

CSV is plain text, so any editor works. Type your data with commas between fields and newlines between rows, save with a .csv extension:

name,role,hire_date
Ada Lovelace,engineer,1843-10-12
Grace Hopper,admiral,1944-07-02
Katherine Johnson,mathematician,1961-04-01

If any field contains a comma, a newline, or a double quote, wrap the whole field in double quotes and double up any internal quotes:

name,quote
Ada Lovelace,"She said, ""hi"""
Grace Hopper,"known for COBOL, FLOW-MATIC"

Saving in a UTF-8-aware editor (VS Code, Sublime, Notepad on Windows 10+) keeps accented characters intact. Good for small files and for debugging an existing CSV; brittle for anything more than about 20 rows.

3. From code (Python, Node, anywhere)

When the data lives in a program, write it straight to CSV from there:

Python:

import csv

rows = [
    {"name": "Ada Lovelace", "role": "engineer"},
    {"name": "Grace Hopper", "role": "admiral"},
]

with open("out.csv", "w", newline="", encoding="utf-8") as f:
    writer = csv.DictWriter(f, fieldnames=["name", "role"])
    writer.writeheader()
    writer.writerows(rows)

Node.js (no library):

import { writeFileSync } from "node:fs";

const rows = [
  { name: "Ada Lovelace", role: "engineer" },
  { name: "Grace Hopper", role: "admiral" },
];

const headers = Object.keys(rows[0]);
const escape = (v) =>
  /[",\n]/.test(v) ? `"${String(v).replace(/"/g, '""')}"` : v;

const csv = [
  headers.join(","),
  ...rows.map((r) => headers.map((h) => escape(r[h])).join(",")),
].join("\n");

writeFileSync("out.csv", csv, "utf-8");

Use this when the data comes from a database query, an API response, or a generation script. It scales to any size because you can stream row-by-row without holding everything in memory.

4. From JSON (with a converter)

When the data starts as JSON — API response, config file, log output — the fastest path is a direct converter. Paste the JSON into our JSON to CSV tool, adjust delimiter and flatten-nested settings, download. It handles nested objects (dot notation), arrays, and quoting automatically. For Excel-specific output, use the JSON to Excel version which pre-selects UTF-8 BOM and CRLF line endings.

On the command line, jq handles this too: jq -r '(.[0] | keys_unsorted) as $k | $k, (.[] | [.[$k[]]]) | @csv'.

Three gotchas that bite everyone once

1. Missing UTF-8 BOM and Excel mangles accents. If the file will be opened in Excel, the file needs to start with the three bytes EF BB BF. Writing from Python? open('out.csv', 'w', encoding='utf-8-sig') adds the BOM. From our tool, toggle "Excel BOM" on.
2. Forgetting to escape commas inside fields. A name like "Smith, John" written unquoted breaks the column count. Always quote fields that contain the delimiter.
3. CRLF vs LF on Windows. Excel on Windows expects CRLF. LF-only files sometimes work and sometimes get read as one giant row. When in doubt, use CRLF.

FAQ