Run any LLM,
learn how
it works.
Free tools to size and rent the GPUs your model needs, with hourly prices from11 providers (AWS, Google Cloud, Azure…). Plus hands-on lessons on gradient descent, MCP, compilers and data lakehouses, all running in your browser.
7 experiences
Tools & lessons.
Calculators you can use today and hands-on lessons on AI and data engineering. Everything runs in your browser, nothing to install.
- Tool
LLM VRAM Calculator
How many GPUs a model needs, with the real KV cache of 2026 architectures. Pick a recent model, new releases are added automatically, or paste any Hugging Face link.
Auto-updated models · quantization · speed & cost · deploy command
- Tool
GPU Rental Prices
The cheapest on-demand price for each GPU across 11 providers (AWS, Google Cloud, Azure…), checked every hour.
H100 · H200 · B200 · RTX 4090 · daily history
- Course
Gradient Descent
Watch optimization happen in 2D and 3D, tune the learning rate, then train a small neural network.
2D & 3D · learning rate · neural network
- Course
Model Context Protocol
What MCP is, how clients and servers talk over JSON-RPC, and how to write your own server.
Architecture · JSON-RPC · servers
- Course
AI/ML Architect
Design AI systems at scale: data, distributed training, inference, MLOps and cost trade-offs.
Training · inference · MLOps
- Tutorial
Build a Lakehouse
What makes a good lakehouse, then deploy one on Kubernetes, step by step.
Kubernetes · MinIO · Trino · Iceberg
- Tutorial
Build Your Compiler
Write a compiler for a small language, from lexing to assembly, until it compiles itself.
S0 → S5 · parsing · assembly · self-hosting