Building JSON APIs in Flask


By the end of this chapter, you should be able to:

  • Use flask-restful to simplify API creation
  • Build an API that performs CRUD on a single resource

Using flask-restful

If you’re building an API that conforms to the conventions of RESTful routing, there’s a module you can use to help simplify the process, called flask-restful. Flask restful allows us to isolate our API logic in a cleaner fashion and makes it easier to add associations with our API. For a simple working example using Flask restful just for the routing (without worrying about a database), check out the docs. For now, we’ll focus on refactoring our existing Book app using flask-restful.

First, we’ll need to install it:

pip install flask-restful

Next, we’ll need to import the things we need from it. We’ll need to use this module in our project/books/ file, so let’s head to the top of that file and modify what we’re importing:

from flask import Blueprint, request, jsonify
from flask_restful import Api, Resource
from project.books.models import Book
from project import db

To create an API out of our blueprint, we’ll need to pass our blueprint into the Api class. We can rename our books_blueprint and do the following:

books_api_bp = Blueprint(

books_api = Api(books_api_bp)

Now that we have an Api, we can start adding resources to it. The syntax here is as follows:

@some_api.resource('/path') # use our api as a decorator!
class SomeAPI(Resource): # create a class that inherits from Resource
    def get(self): # methods inside the class should correspond to HTTP verbs

    def post(self):

In our case, this means we can refactor our previous code to the following:

class BooksAPI(Resource):
  def get(self):
    return jsonify([
      {'id':, 'author':, 'title': book.title}
      for book in Book.query.all()

  def post(self):
    new_book = Book(
    return jsonify({
      'title': new_book.title

Finally, we’ll need to update our to import our Api blueprint:

from project.books.views import books_api_bp

app.register_blueprint(books_api_bp, url_prefix='/api')


This code is better, but there’s still a fair amount of redundancy in our API methods. For example, all of the methods in our BookApi have to manually serialize data and pass it into jsonify. Fortunately, Flask Restful also comes with serialization functionality, so that we can more easily convert between objects and JSON data.

Here’s how it works: we need to use Flask-Restful’s marshal_with function to specify how we should marshal our data. (You can think of marshaling and serializing as essentially synonymous. If you’d like to get into the weeds about the distinction, Stack Overflow has got your back.) To help with the marshaling, we’ll also need to import something else from Flask-Restful. 

At the top of your project/books/, you can include the following:

from flask_restful import Api, Resource, fields, marshal_with

book_fields = {
  'id': fields.Integer,
  'title': fields.String,
  'author': fields.String

You can then marshal your object data using the book_fields dictionary, which tells your application how to treat different attributes in your objects. To actually marshal the data, you use the @marshal_fields decorator. Here’s how the get inside of the BookListApi can be refactored:

def get(self):
    return Book.query.all()

Even though it looks like we’re just returning a list of Book instances, including the decorator ensures that our objects will eventually be converted properly, so that the server can respond with JSON.


  1. Refactor the rest of your books app using what you’ve learned about flask_restful.
  2. Build a CRUD app on a single resource that uses AJAX exclusively: no page reloads allowed.

When you’re ready, move on to Authenticating a Flask API with JWTs

Download the
program syllabus

Get the syllabus, daily schedule, and a few more details about Rithm: