Google's Machine Learning Crash Course is one AI course that expects you to know some Python. Its prerequisites page puts it plainly: "Ideally, you should have some experience programming in Python because the programming exercises are in Python." Our reading, based on that page: if you want to learn AI hands-on, a short Python course is a sensible first step.
This list has 5 free Python courses. Each one's own page says it is free or no cost, and we checked each page in November 2026. For each pick you get who it is for, what it covers, how long the page says it takes, and what it says about certificates. We also list one reference that Google points to, and the Python topics Google says you need before its machine learning course.
If you are new to AI as a whole, our 12-week plan to learn AI from scratch shows where Python fits. This page is only about the Python step.
The 5 picks at a glance
- Python for Everybody (PY4E): for people with no programming background. Free materials, lectures, book and assignments.
- CS50's Introduction to Programming with Python (CS50P): for people with or without programming experience. Ten weeks of material, free through Harvard's OpenCourseWare.
- Kaggle Learn: Python: a short course of 7 lessons, estimated at 5 hours. No cost.
- Google's Python Class: for people with a little programming experience. Free written materials, lecture videos and exercises.
- Kaggle Learn: Pandas: the data-handling next step, estimated at 4 hours. No cost.
If you have never coded
Python for Everybody (PY4E)
Python for Everybody is a course site by Charles R. Severance. The home page says: "This web site is building a set of free materials, lectures, book and assignments to help students learn how to program in Python." The lessons page opens: "Hello and welcome to my site where you learn Python even if you have no programming background."
- Who it is for and prerequisites: the lessons page says you do not need a programming background.
- What it covers: the lessons page starts with installing Python and moves through variables, conditional execution, functions, loops, strings, files, lists, dictionaries, tuples and regular expressions.
- What is free: the home page says that if you log in, you have joined a free, global open and online course with a grade book, autograded assignments, discussion forums and badges.
- Certificates: the home page offers free certificates for University of Michigan students and staff. It also links to versions of the course on Coursera, edX and freeCodeCamp. We did not check the terms on those sites, so this list does not say what they cost.
CS50's Introduction to Programming with Python (CS50P)
CS50P is Harvard's Python course, taught by David J. Malan. The page says it is designed for students with or without prior programming experience who would like to learn Python specifically.
- What it covers: functions, variables and types, conditionals, loops, exceptions, debugging, unit tests, third-party libraries, regular expressions, classes and objects, and reading and writing files.
- Length: ten weeks of material, each with a lecture and a problem set, then a final project.
- What is free: the page says: "Even if you are not a student at Harvard, you are welcome to “take” this course for free via this OpenCourseWare by working your way through the course’s ten weeks of material."
- What you need: the page says no software is required except a web browser. To submit problem sets and the final project for feedback, it asks you to create an edX account.
- Certificates: for a verified certificate or a professional certificate from edX, the page tells you to enroll through cs50.edx.org instead. The CS50P page does not give a price for those.
If you want a short course
Kaggle Learn: Python
Kaggle's Python course has 7 lessons, each with an exercise: Hello, Python; Functions and Getting Help; Booleans and Conditionals; Lists; Loops and List Comprehensions; Strings and Dictionaries; and Working with External Libraries.
- Length: the page shows an estimate of 5 hours.
- What is free: the page lists the cost as "No cost, like all Kaggle Learn Courses."
- Why it matters for AI: list comprehensions and external libraries both appear on Google's list of Python topics for its machine learning course (see below). Kaggle covers both in this course.
If you already code in another language
Google's Python Class
Google's Python Class says: "Welcome to Google's Python Class -- this is a free class for people with a little bit of programming experience who want to learn Python." It includes written materials, lecture videos and code exercises.
- Who it is for: the page says it is for people with enough programming experience to know what a variable or an if statement is.
- What it covers: the first exercises work on strings and lists, and the later exercises are full programs that deal with text files, processes and http connections.
- Length: the page says that at Google, the material makes up an intensive 2-day class.
- Note: the page was last updated on July 23, 2024, according to its footer.
Once you know the basics
Kaggle Learn: Pandas
Pandas is one of the two Python libraries Google lists as prework for its machine learning course, next to NumPy. Kaggle's Pandas course has 6 lessons: Creating, Reading and Writing; Indexing, Selecting & Assigning; Summary Functions and Maps; Grouping and Sorting; Data Types and Missing Values; and Renaming and Combining.
- Length: the page shows an estimate of 4 hours.
- What is free: the page lists the cost as "No cost, like all Kaggle Learn Courses."
The Python you need for Google's ML Crash Course
Google's Machine Learning Crash Course is one next step once you know Python. Its prerequisites page lists the Python basics it expects and says they are covered in The Python Tutorial:
- Defining and calling functions, with positional and keyword parameters.
- Dictionaries, lists and sets: creating, accessing and iterating.
- For loops, including for loops with multiple iterator variables.
- If/else conditional blocks and conditional expressions.
- String formatting.
- Variables, assignment and the basic data types int, float, bool and str.
- List comprehensions, which Google says a few of the exercises use.
The same page asks learners new to NumPy or pandas to do Google's NumPy and pandas UltraQuick Tutorial Colab exercises first. It says the programming exercises run in your browser using the Colaboratory platform, with no setup required. For more on free notebooks and GPUs, see our guide to free GPU access with Colab, Kaggle and student credits.
On the course's own page, Google describes Machine Learning Crash Course as a fast-paced, practical introduction to machine learning, with animated videos, interactive visualizations and hands-on practice exercises. We did not find a statement about price on that page, so this list does not call it free.
A reference, not a course: The Python Tutorial
The Python Tutorial is part of the official Python documentation, and it is the reference Google's prerequisites page points to. Its own tip says: "This tutorial is designed for programmers that are new to the Python language, not beginners who are new to programming." The tutorial page does not describe itself as a course or mention a price, so we list it separately from the 5 free picks.
How to pick
- No coding at all: Python for Everybody says it is for people with no programming background. CS50P says it is for students with or without prior experience.
- A few hours to spare: Kaggle's Python course is estimated at 5 hours. It is the only Python-basics pick here whose page gives an estimate in hours.
- You code in another language: Google's Python Class is written for people with a little programming experience.
- You know Python and want AI: Kaggle's Pandas course, then Google's prework list and Machine Learning Crash Course.
Our reading: whichever course you take, check it against Google's list above. If you can do everything on it, you have the Python that Google's machine learning course expects.
Where to go next
- Our list of free AI courses and our free LLM courses for the AI step.
- Math for machine learning for the algebra and statistics Google's prerequisites page lists.
- How to become a data scientist and how to become an AI engineer for career paths that use Python.
- How to build an AI portfolio once you have projects to show.
Bottom line
Google's Machine Learning Crash Course says its exercises are in Python, and its prerequisites page lists the Python you need. The 5 courses above each say on their own page that they are free or no cost. PY4E and CS50P say they work for people with no programming experience. Kaggle's Python course has a 5-hour estimate. Google's Python Class is written for people with a little programming experience. Kaggle's Pandas course covers the data step. Check certificate terms on the provider's page before you pay for anything.
