How to Pretty Print JSON in Python (3 Methods)

Published 2026-06-16 · DevFormat AI Guides

Python's standard library covers JSON pretty-printing completely — no packages needed. Here are the three methods, from the everyday one-liner to command-line piping, plus the gotchas that catch people.

Method 1: json.dumps with indent

import json

data = {"name": "Amina", "roles": ["admin", "editor"], "active": True}
print(json.dumps(data, indent=2))

Output:

{
  "name": "Amina",
  "roles": [
    "admin",
    "editor"
  ],
  "active": true
}

Useful parameters:

• indent=2 — spaces per level (use indent="\t" for tabs).
• sort_keys=True — alphabetical key order, which makes diffs between two payloads meaningful.
• ensure_ascii=False — output real Unicode characters instead of \u00e9 escapes. Almost always what you want for human-readable output.

Method 2: reading and writing files

import json

with open("config.json") as f:
    data = json.load(f)

with open("config.pretty.json", "w", encoding="utf-8") as f:
    json.dump(data, f, indent=2, ensure_ascii=False)

Note the pairing: load/dump work with file objects, loads/dumps with strings. Mixing them up is the classic beginner error, and the resulting message (expected str, bytes or os.PathLike) does not obviously point at it.

Method 3: the command line — no script at all

Python ships a JSON pretty-printer as a module, perfect for piping:

# pretty print a file
python -m json.tool data.json

# pretty print an API response
curl -s https://api.example.com/users | python -m json.tool

# with sorted keys (Python 3.5+)
python -m json.tool --sort-keys data.json

This doubles as a validator: invalid JSON exits with an error and the line number. If you have jq installed, jq . does the same with colour — but json.tool is already on any machine with Python.

What about pprint?

from pprint import pprint
pprint(data)

pprint formats Python objects, not JSON: you get single quotes, True instead of true, and None instead of null. Fine for debugging your own dicts; wrong the moment the output needs to be JSON that anything else parses. If the result will be consumed, always use json.dumps.

Common gotchas

• Datetimes are not serialisable — json.dumps(datetime.now()) raises TypeError. Convert with default=str or serialise ISO strings explicitly.
• Dict keys become strings — integer keys are silently converted, so a round trip does not return the identical structure.
• NaN sneaks through — Python happily emits NaN, which is invalid JSON that other parsers reject. Pass allow_nan=False to catch it at source.

And when you have a payload in your clipboard rather than in code, skip Python entirely: paste it into our free JSON formatter — it pretty-prints, validates with line numbers, and never uploads your data.