Loading…
406 chapters · 3 courses · updated every hour
Real courses on AI agents, machine learning and robotics, with code you can run, quizzes that make it stick, a tutor beside you, and a live feed of everything that changes in AI.
One-click sign-in with Google. We only receive your name and email.
Why Nanofuze
Not a video playlist and not a wall of links: one place that teaches, tests you, keeps you current and answers your questions.
Every example marked as runnable was executed, and the output printed under it is the output you get. No hand-waving, no code that only looks right.
Questions inside the chapters feed a spaced-review queue that brings an idea back just before you would forget it.
Nano knows the chapter you are on. Ask it to explain simply, quiz you or set an exercise, by typing or by voice.
6 week-by-week programs of 50 hours or more, each with a hands-on lab every week: robotics, AI from zero, agents, machine learning, modern AI models and building projects.
A model finder, a cost calculator and a can-I-run-it-locally check, built on live prices and rankings. Results live in the link, so you can share them.
One page shows what changed in AI: new models, price moves, leaderboard shifts, releases and rules, each with its numbers and where they came from.
The courses
Each course is a complete syllabus from the first idea to a working project, with checks along the way.
92 chapters · 55.1 hours
Learn AI from the basics and build agents: tools, MCP, retrieval, hooks and teams.
481 code samples72 quizzes13 parts
View the syllabus
159 chapters · 145.7 hours
From a line through points to large language models: the maths, classical models, neural networks, vision, language, generative AI and production, built from scratch.
505 code samples308 quizzes18 parts
View the syllabus
155 chapters · 118.8 hours
From kinematics to humanoids: control, perception, learning, vision-language-action models, ROS 2, safety and projects.
586 code samples120 quizzes16 parts
View the syllabus
Learning paths
A path is a week-by-week program to one goal, 50 hours or more, with a hands-on lab every week. Pick the one that matches what you want to build or understand.
Curious about robots and embodied AI
A 24-week program: 13 weeks from coordinate frames to learned robot policies, with 13 labs on any laptop, then an 11-week build of NanoRover, a real robot with vision, hearing, many sensors and a safe AI brain, from parts list to wiring to code.
24 weeks123 h123 chapters
Start hereNo background needed
A 13-week program for a complete beginner: Python and data, your first models, how neural networks and language models work, and your first AI agent, with a hands-on lab every week.
13 weeks64 h75 chapters
Developers
A 14-week program from the agent loop to a launched agent: tools, memory and retrieval (RAG), MCP, the Claude and OpenAI SDKs, many agents, evals, safety and shipping to production, with hands-on labs every week and a capstone project.
14 weeks60 h92 chapters
Aspiring ML engineers and analysts
A 14-week program: the maths in a few pages, honest data work, classical models and neural networks you build from scratch, then PyTorch, MLOps and serving, with a lab every week and an end-to-end capstone.
14 weeks67 h91 chapters
You use them; now see inside
A 14-week program on how today's models work: convolutions, tokenisers and attention, a small GPT you train, adapters, quantisation and evaluation, then autoencoders, diffusion and reinforcement learning, with a lab every week.
14 weeks70 h69 chapters
Hands-on, code first
A ladder of twelve projects that grow in ambition, each with the lab behind it: a spam filter, house prices, a digit recogniser, an image classifier, a recommender, a language model, three agents, a simulated robot arm and an ML capstone.
12 weeks72 h24 chapters
Always current
Every hour we read public data (model prices, the human-preference leaderboard, Hugging Face, software releases and government documents) and show what changed. No rewritten articles, no opinions: numbers, and a link to where each one came from.
DeepSeek V4 Pro 0813 output price raised 102%: $1.98 → $4 per 1M tokens
DeepSeek · input and output prices both changed · 1.05M context · 4 h ago
GLM 4.7 is scheduled to shut down on Dec 31, 2026
Z.ai · listed to be removed from OpenRouter in 60 days · 4 h ago
GLM 4.7 output price raised 11%: $1.98 → $2.2 per 1M tokens
Z.ai · input and output prices both changed · 205K context · 4 h ago
Transformers 5.22.0 is out
A new minor version of the library most open models are published for · 2 days ago
Intrinsic Distribution Matching
234 upvotes on Hugging Face Daily Papers · 10 authors · posted Oct 29, 2026 · 3 days ago
115 more, with prices, the leaderboard and new models, on the news page.
The Lab
Small, honest calculators for the decisions every AI builder makes. Everything you enter stays in your browser, and every result shows how it was worked out.
Choose
Describe the job (coding, writing, vision, long documents) and see the models that rank best for it, with their price and context size.
Compare
Enter your volume and compare what the same workload costs on different models, from live price lists.
Check
Pick a model and your hardware, and see whether it fits, how much memory it needs and roughly how fast it will run.
How you learn
01
Short chapters written to be understood on the first read: the idea, a picture, an example and what to remember.
02
Run the examples and see real output. Change a line, run it again, and watch what happens.
03
Answer the questions inside the chapter. What you miss comes back for review, spaced out so it sticks.
04
Stuck? Nano explains it another way, quizzes you or sets an exercise, without leaving the page.
Who it is for
If you learn best by doing, this is for you.
You can code. Learn how agents, tools, retrieval and evaluation fit together, and ship something that works.
The agents path
Start from zero with plain language and pictures, and build up to real models without skipping the maths that matters.
Start from zero
Know what is possible, what it costs and what to ask your team, with sources you can check.
Modern AI models
Go from sensors and control to learning-based robots, with build projects at the end of each part.
The robotics path
Built to be trusted
A learning site is only worth your time if you can believe it.
News items show their numbers and link to the data they came from. Nothing is copied or rewritten from other sites.
How important a change is comes from written, public rules, not from a language model's opinion.
Each course says what was run and what was only checked, and flags facts that come from a public source we could not verify against the original.
On the learning site we receive your name and email from Google, nothing else. Your progress and notes stay in your own browser.
Questions
A free place to learn AI by building it. The learning part lives on its own site, learn.nanofuze.com: learning paths (goal-shaped routes such as AI from zero, building agents, how large language models work, robotics and physical AI), three full courses with code that runs, quizzes and spaced review, Here on nanofuze.com you find a Lab of tools, a glossary, and one news page that keeps you up to date with new models, prices, research and rules.
It is free. News, the Lab, the glossary and the rest of nanofuze.com need no account. The learning site (learn.nanofuze.com) asks you to sign in with your Google account; it receives your name and email address and nothing else. Your progress, notes, daily goal and review queue are saved in your own browser, not on your account, and are never sent to us. Clearing your browser's site data starts you again from the beginning.
If you are new to AI, take the learning path called AI from zero: seventeen short chapters that explain what AI, machine learning and language models are and how to use them well. If you can already code and want to build with language models, take Build AI agents. The learning site has a line for each kind of reader.
A course teaches one field in its own order, from the first idea to the last. A learning path picks the chapters you need, from any course, for one goal (for example, how large language models work) and groups them in stages, so you always know what to read next.
Each chapter is short, usually three to ten minutes. Code examples run, and the real output is shown beneath them. Quizzes inside chapters and questions at the end make you recall answers, and every question you answer joins a review queue that brings it back just before you would forget it. Explorables let you drag a slider and watch an idea change, and Nano, the assistant on the learning site, knows which chapter you are on and can explain it, quiz you or set an exercise.
Yes. Every Python example marked as runnable was executed with the libraries named on the course page, and the output shown is what it printed. A few short snippets, such as one that downloads pretrained weights, were only checked for syntax and are not marked as runnable. Each course says what was run and what was not.
Sign in with Google and start with the path or course that fits you. Your progress is saved as you go.