How to prepare for the IOAI

The International Olympiad in Artificial Intelligence (IOAI) is the most selective AI competition for high-school students, with national teams from more than 60 countries. Making the team requires real machine-learning skills: Python, data manipulation, classical models, neural networks, computer vision and natural language processing. This platform was built by IOAI coaches to train for exactly that, and it is free.

588
students
89
countries
8,330
exercises solved

A method built for mastery, not for watching videos

You learn by solving problems, in the spirit of the olympiad itself. Every concept is introduced through a concrete problem that you solve in the browser; an automatic judge runs your code against hidden tests and tells you honestly where you stand. Solutions unlock only once you have solved the exercise, hints are not handed out, and every student pledges not to use generative AI on the exercises: what you measure here is what you actually know.

A six-level path, from Python to deep learning and beyond

The curriculum is organised like France-IOI's famous training tower: six levels, each closed by a boss challenge calibrated on real olympiad difficulty. You start with Python, NumPy and data wrangling, move through classical machine learning (regressions, trees, clustering, evaluation methodology), then neural networks, computer vision and NLP, up to transformer-based models; a sixth, frontier level goes on to ML systems, post-training and research papers. Chapter by chapter:

Level 1

Python: the basics · NumPy & vectors · Pandas & visualization · Probability · Beginner algorithms (optional)

Level 2

Statistics · Supervised models 1 · Supervised models 2 · Introduction to unsupervised learning · Evaluation methodology · Data preparation & feature engineering · Complete pipeline · How a machine learns

Level 3

Linear algebra for ML · Computing a derivative (bonus) · Optimization · Loss functions & information theory · Neural networks · Backpropagation · Training with PyTorch · Intermediate algorithms (optional)

Level 4

Vision · Text & embeddings · Sequences & encoder-decoder · Audio · Finetuning 1 · Few labels, shifting data · Mixtures, anomalies, matching

Level 5

Transformers · Training a language model · Finetuning 2 · LLMs in practice · Advanced vision · Image generation · Olympiad problems · AI competition problems · Retrieval and representations · Reinforcement learning

Level 6

Inference & ML systems · Foundations: compression & complexity · Advanced architectures · Post-training · Foundational readings

A worldwide community

Students from 89 countries train here: 588 so far, with 8,330 exercises solved. Daily questions, revision dojo, duels between friends and a help forum where solvers coach the ones still stuck: the competition spirit, without the solitude.

Frequently asked questions

Is the platform free?

Yes. Training is free for students: create an account and start. National coaches can request a follow-up dashboard for their team.

What level do I need to start?

None. The curriculum starts at Python basics and a 2-minute placement test suggests your starting level; strong students can skip ahead by beating a level's boss challenge.

How is it different from video courses?

You learn by solving problems: an automatic judge grades your code on hidden tests, solutions unlock only after you solve each exercise, and every student pledges not to use generative AI on the exercises.

Can national teams use it to prepare for the IOAI?

Yes. The platform is used by students from 75+ countries, is available in English and French, and offers national coaches a dashboard to follow their team (contact@artix.tech).

Start today

Registration is free and takes a minute. A placement test suggests where to start; national coordinators can follow their team's progression.

Create a free account