Reading & Writing JSON in Python

Python provides the built-in json module (no installation required).

There are 4 core functions you must master. Here is the easiest memory trick:

Function	The Letter 's' Meaning	Where Data Comes From / Goes To
json.loads()	Load String	Reads from a JSON text string in memory into a Python dict
json.dumps()	Dump String	Converts a Python dict into a JSON text string
json.load()	No 's' (File)	Reads from a JSON file on disk into a Python dict
json.dump()	No 's' (File)	Writes a Python dict into a JSON file on disk

---

1. Working with Strings in Memory: loads() & dumps()

Converting JSON String → Python Dictionary (loads)

When you receive JSON text from an API response, use json.loads() to convert it to a usable dictionary:

import json

# JSON text received from a web server
api_response_text = '{"name": "Sai Kumar", "role": "AI Engineer", "active": true}'

# Deserialize: String -> Python Dictionary
user_data = json.loads(api_response_text)

print(type(user_data))         # Output: <class 'dict'>
print(user_data["role"])       # Output: AI Engineer
print(user_data["active"])     # Output: True (converted to Python bool)

---

Converting Python Dictionary → JSON String (dumps)

When you want to send data to an API or print clean JSON text, use json.dumps():

import json

student = {
    "name": "Kavya",
    "marks": 94,
    "passed": True
}

# Serialize: Python Dictionary -> JSON String
json_string = json.dumps(student, indent=2)

print(type(json_string))  # Output: <class 'str'>
print(json_string)

Output:

{
  "name": "Kavya",
  "marks": 94,
  "passed": true
}

Why indent=2?

By default, json.dumps() produces one compact continuous line. Adding indent=2 or indent=4 formats the output with beautiful spacing (pretty-printing).

---

2. Working with Files on Disk: load() & dump()

Writing Data into a .json File (json.dump)

Use json.dump() with a with open(...) file handle:

import json

app_config = {
    "app_name": "Think IT Telugu LMS",
    "version": "2.4.0",
    "theme": "dark",
    "max_upload_mb": 50
}

# Save configuration directly into settings.json
with open("settings.json", "w") as file:
    json.dump(app_config, file, indent=4)

print("✅ settings.json file saved successfully!")

---

Reading Data from a .json File (json.load)

Use json.load() to read an existing JSON file directly into a Python dictionary:

import json

with open("settings.json", "r") as file:
    loaded_config = json.load(file)

print("Loaded App Name:", loaded_config["app_name"])
print("Active Theme:", loaded_config["theme"])

---

3. Common Beginner Mistake: load vs loads

The Extra 's' Matters!

• json.loads(file) ❌ Crashes! (loads expects a text string, not a file object).
• json.load(file) ✅ Correct for files.
• json.load('{"id": 1}') ❌ Crashes! (load expects an open file handle, not a string).
• json.loads('{"id": 1}') ✅ Correct for strings.

---

Quick Summary

• json.loads(str): Deserializes a JSON string into Python dict/list.
• json.dumps(obj): Serializes a Python dict/list into a JSON formatted string (indent=4 for pretty-printing).
• json.load(file): Reads and parses JSON directly from an open file handle.
• json.dump(obj, file): Writes a Python data structure directly to an open disk file.
• Memory vs. File: The s suffix stands for String (loads/dumps), while methods without s operate on Files (load/dump).

---

What's Next?

Let's explore why LLMs, AI agents, and OpenAI/Anthropic APIs rely fundamentally on JSON in Why AI Uses JSON!