If you use Pylance in VS Code for its completions, navigation, or GitHub Copilot integration, you can now use Pyrefly as its type checking backend. We have been working with the Microsoft Python team to make this possible. The result is a new diagnostics source option in Pylance that uses Pyrefly instead of Pyright to power the type analysis backend, keeping the Pylance experience intact.

Pyrefly v1.4 is out. Highlights:

• Fewer false positives, from fixes to import resolution, gradual types, and overloads.
• Tensor shapes for libraries: a new compatibility syntax, and JAX shape stubs on PyPI.
• The VS Code extension uses the Pyrefly in your environment, so your editor and CLI run the same version.
• uv workspace members are now type checked.
• Layered configuration and BasedPyright migration.

A question that has been on our minds lately is how to build a community around Pyrefly when AI is rapidly reshaping longstanding open source norms. I don't think anyone has fully figured this out yet, including us, but we're making some changes to Pyrefly's contributing guidelines that we think are a good start. Here's what we've seen change in open source and what we're doing about it.

We've just released Pyrefly v1.3, which adds flexible tools for configuring, suppressing, and baselining errors. Also included: a composable tensor shape DSL that expands experimental tensor shape support for JAX and NumPy, improved DataFrame schema checking for Polars, and a battery of new type checking diagnostics.

Pyrefly is one of the major new Python type-checkers taking the world by storm. It's often touted for its speed, usability, and typing-spec-conformance. But it also offers an extra nifty little feature you may never have heard of: type-coverage.

Let's learn about what it is, why you might want to care about it, and what it can do for you.

Pyrefly v1.2 is now available 🎉 This release adds support for Bazel and attrs, along with improvements to type checking, the language server, and the CLI.

Companion post to my PyCon 2026 Typing Summit talk, "From Soundness to Blame: AI-Assisted Formalization of a Tiny Python." This is the first in a series.

Built-in support for attrs is available in Pyrefly as of version 1.2.0-dev.1, and ships in the upcoming 1.2.0 stable release. You get accurate constructor signatures and better type safety for your attrs classes, with no plugins or configuration to set up.

Pyrefly v1.1 is now available 🎉 This is our first minor release since reaching stable v1.0 in May, and it brings improvements across the board: better performance, smarter type checking, and new IDE refactoring tools.

Why aren't tensor shapes part of the type system?

When you write a PyTorch model, the hardest part of composing tensor ops is keeping track of their shapes. The standard practice today is to write shapes down as comments as you go. This talk introduces an experimental Pyrefly feature that brings tensor shapes into Python's type system, so those shape annotations can become inferred type hints instead of comments.

The talk was originally presented at the PyCon US 2026 Typing Summit. The slides and edited transcript are provided below for your convenience.

📎 Slides (PDF)

Want to give it a spin? This feature is available in Pyrefly today, and we'd love to hear your feedback and suggestions.