Object detection — the AI puts a box around each thing and says how sure it is.
That AI is real — no video, no trick. It's looking at the photo right now, the same way a self-driving car does. Computer science, maths and AI from the ground up, in Cases you solve with real Python running in the browser. No installs, any laptop, any age.
No sign-up needed to try. Level 1 is free, forever.
🚀 AI Frontier · the AI everyone's talking about
No prerequisites, no waiting — every learner begins here. Run real Python in the browser and see how today's AI sees, reads, writes, connects ideas and takes action.
Upload a photo and watch a real AI box what it sees — live, in your browser.
Teach a computer to read, and judge the mood of a message.
Build a tiny next-word chatbot and find its creativity dial.
Connect the dots — how everything links to everything.
The newest AI that senses, decides and takes action.
Real tools · not toys
You write real code with the actual industry libraries — all running in the browser, no installs, on any laptop. The badges below are what you actually type.
The world's #1 language for AI and data.
Runs Python in your browser — zero installs.
Fast maths on a whole row of numbers.
Spreadsheets you can query with code.
Turn numbers into charts you can see.
Train real models — the classic ML toolkit.
Real neural networks, right in the browser.
A pretrained model that spots 80 objects.
From a first chart to a real neural network — see how it all fits together on the curriculum map.
A real problem worth solving — “Is this dataset telling the truth?”
Meet the libraries. NumPy = boxes of numbers. Matplotlib = crayons for charts.
Write and run real Python, right in the browser. No installs.
A chart appears and you see the answer. Seeing beats being told.
A twist to crack on your own — now the skill is yours.
A chart is a picture of numbers. Write your first lines of Python and see them draw.
Python · Matplotlib
Typically 7 and under — or any adult starting cold
Do maths on a whole row of numbers at once, the way real data work is done.
+ NumPy
Typically 8–11
Load real datasets, ask them questions, and defend the answer.
+ Pandas
Typically 12–15
Train real models and understand why they predict what they predict.
+ scikit-learn
Typically 16+
Real code, a real chart, a real answer — no setup, no credit card, no experience.
▶ Open the free Case