In this tutorial we’ll be converting a JSON file to CSV using Python.

Let’s say we have a JSON file that looks like this:

[
    {
        "id": 1,
        "name": "Albert",
        "address": {
            "city": "Amsterdam"
        }
    },
    {
        "id": 2,
        "name": "Adam",
        "address": {
            "city": "Paris"
        }
    },
    {
        "id": 3,
        "name": "Sara",
        "address": {
            "city": "Madrid"
        }
    }
]

Step 1: Install pandas

You could technically use the standard json and csv modules from Python to read the JSON file and write a CSV file, but depending on how your JSON file is structured, it can get complicated, and pandas has some great functions to handle these cases, without mentioning it’s really fast, so that’s what we will use.

You can install pandas using pip running this command:

$ pip install pandas

Step 2: Load the JSON file

Let’s load the JSON file using the json Python module:

import json

with open('input.json', encoding='utf-8') as file:
    data = json.loads(file.read())

Step 3: Normalize the JSON data

Because the JSON format can hold structured data like nested objects and arrays, we have to normalize this data so that it can be respresented in the CSV format.

A quick way to do this is to use the pandas.normalize_json function, which will take JSON data and normalize it into a tabular format, you can read more about this function here.

import pandas

df = pandas.json_normalize(data)

In our case, if we print the resulting dataframe, it will look something like this:

print(df)

   id    name address.city
0   1  Albert    Amsterdam
1   2    Adam        Paris
2   3    Sara       Madrid

As you can see the address which was an object, was flattened out to a column.

Step 4: Export to CSV

Now that our data is normalized to a tabular format, let’s write our CSV file!

Pandas has a convenient function for this:

df.to_csv('input.csv', index=False, encoding='utf-8')

And here is the resulting CSV file:

id,name,address.city
1,Albert,Amsterdam
2,Adam,Paris
3,Sara,Madrid

If you want to configure the output, check out the documentation page for the to_csv function here

Putting everything together

Here’s the final code to convert JSON to CSV:

import json
import pandas


with open('input.json', encoding='utf-8') as file:
    data = json.loads(file.read())

df = pandas.json_normalize(data)

df.to_csv('output.csv', index=False, encoding='utf-8')

Online conversion

If your file is too big or your data is a bit more complex, you can try our online JSON to CSV converter, which can handle huge files and more use cases.