JSON to CSV Converter Online

Example: Nested object

[{"name": "Ada", "address": {"city": "London", "country": "UK"}}]

Becomes: name,address then Ada,"{""city"":""London"",""country"":""UK""}"

Example: Nested array

[{"name": "Ada", "tags": ["developer", "speaker"]}]

Becomes: name,tags then Ada,"[""developer"",""speaker""]"

Browser JavaScript with Papa.unparse

Papa Parse works client-side. Pass an array of objects to Papa.unparse:

import Papa from 'papaparse';

const data = [
  { name: 'Ada', age: 30, city: 'London' },
  { name: 'Bob', age: 25, city: 'Paris' }
];

const csv = Papa.unparse(data);
// "name,age,city\r\nAda,30,London\r\nBob,25,Paris"

Node.js with @json2csv/plainjs

The @json2csv/plainjs package (v6+) handles synchronous conversion; use @json2csv/node for streaming large files:

import { Parser } from '@json2csv/plainjs';

const data = [
  { name: 'Ada', age: 30, city: 'London' },
  { name: 'Bob', age: 25, city: 'Paris' }
];

const parser = new Parser({ fields: ['name', 'age', 'city'] });
const csv = parser.parse(data);

Python with csv.DictWriter

The standard library csv module handles simple cases:

import csv
import io

data = [
    {'name': 'Ada', 'age': 30, 'city': 'London'},
    {'name': 'Bob', 'age': 25, 'city': 'Paris'}
]

output = io.StringIO()
writer = csv.DictWriter(output, fieldnames=['name', 'age', 'city'])
writer.writeheader()
writer.writerows(data)
print(output.getvalue())

Python with pandas (nested JSON)

For nested or complex JSON, use pandas.json_normalize to flatten first:

import pandas as pd

data = [
    {'name': 'Ada', 'address': {'city': 'London', 'zip': 'SW1'}},
    {'name': 'Bob', 'address': {'city': 'Paris', 'zip': '75001'}}
]

df = pd.json_normalize(data)
# Columns: name, address.city, address.zip
df.to_csv('output.csv', index=False)

jq @csv

Pass -r so the row is CSV, not a JSON string of CSV. List the columns; jq does not build a header from object keys unless you emit keys_unsorted first. keys sorts alphabetically. .[0] | keys_unsorted uses only the first object, so a later city key is dropped. An empty array with that header form errors null (null) has no keys:

printf '%s' \
  '[{"name":"John Doe","age":30,"email":"john@example.com","city":"New York"},{"name":"Jane Smith","age":25,"email":"jane@example.com","city":"Los Angeles"},{"name":"Bob Johnson","age":35,"email":"bob@example.com","city":"Chicago"}]' \
  | jq -r '(.[0]|keys_unsorted), (.[] | [.name,.age,.email,.city]) | @csv'
# "name","age","email","city"
# "John Doe",30,"john@example.com","New York"
# "Jane Smith",25,"jane@example.com","Los Angeles"
# "Bob Johnson",35,"bob@example.com","Chicago"

printf '%s' \
  '[{"name":"John Doe","age":30,"email":"john@example.com","city":"New York"}]' \
  | jq '.[] | [.name,.age,.email,.city] | @csv'
# "\"John Doe\",30,\"john@example.com\",\"New York\""

printf '%s' '[{"name":"Ada"}]' | jq -r '.[] | @csv'
# jq: error (at <stdin>:0): object ({"name":"Ada"}) cannot be csv-formatted, only array

printf '%s' '[{"name":"Ada"},{"name":"Bob","city":"Paris"}]' \
  | jq -r '(.[0]|keys_unsorted) as $k | $k, (.[] | [.[ $k[] ]]) | @csv'
# "name"
# "Ada"
# "Bob"

Nested values with tojson

A nested object or array is not a CSV cell. tojson first — the same JSON-in-one-cell shape this page writes with JSON.stringify, not pandas json_normalize dotted columns. jq still quotes Ada; this page leaves that name unquoted:

printf '%s' \
  '[{"name":"Ada","address":{"city":"London","country":"UK"}}]' \
  | jq -r '.[] | [.name,.address] | @csv'
# jq: error (at <stdin>:0): object ({"city":"Lo...) is not valid in a csv row

printf '%s' \
  '[{"name":"Ada","address":{"city":"London","country":"UK"}}]' \
  | jq -r '.[] | [.name, (.address|tojson)] | @csv'
# "Ada","{""city"":""London"",""country"":""UK""}"

printf '%s' \
  '[{"name":"Ada","tags":["developer","speaker"]}]' \
  | jq -r '.[] | [.name, (.tags|tojson)] | @csv'
# "Ada","[""developer"",""speaker""]"