In this tutorial, we explore Pyright, Microsoft’s high-performance static type checker for Python, and walk through its most powerful features in a hands-on, Colab-friendly format. We start from the ground up with basic annotations and type inference, then progressively advance through Union types, type narrowing, generics, Protocols, TypedDict, dataclasses, and modern typing constructs like Self, TypeAlias, and NewType. We also examine how strict mode raises the bar for type safety across an entire codebase and how pyrightconfig.json provides fine-grained control over diagnostic rules at the project level. Additionally, we deliberately introduce both correct and intentionally broken code to see exactly how Pyright catches real-world mistakes before they ever reach runtime.
We begin by installing Pyright and setting up two helper functions, write and run_pyright, which we reuse throughout subsequent sections to create typed Python files and parse Pyright’s JSON diagnostic output cleanly. We then move on to the basics of type annotations, deliberately passing incorrect argument types and mismatched assignments to observe exactly how Pyright flags each violation.