pytorch / pytorch/executorch

Official Windows Support

Open
#13,781 0 comments 1 reaction 1 assignee View on GitHub

@GregoryComer is already working on this.

Since Aug 28, 2025.

module: build/install release notes: windows
Dominant language
Python
Stars
5k
Forks
1.2k
Avg merge
2d 10h
Merged PRs (30d)
581

Description

Summary

This is the top-level tracking task for official Windows support. Native Windows support is in experimental status, meaning that the build interface, feature set, and timeline are subject to change.

Features

Support is planned for both the ahead of time (Python) and runtime on native Windows. The following are targeted for official support:

  • Python Packages: Install the ExecuTorch python package through pip or install from source using install_executorch.bat.
  • Runtime Build: Compile the C++ runtime for x86 and x84-64 using Clang-CL.
  • Python Runtime Bindings: Run models from Python.
  • Kernel Libraries: Build the portable, optimized, and quantized operator libraries.
  • XNNPACK Backend: Build the XNNPACK (CPU) backend.

Disclaimers

  • Building using MSVC is not supported. You'll need Clang-CL. See Clang/LLVM support in Visual Studio for installation instructions.
  • Backends other than XNNPACK aren't officially supported on Windows (yet).
  • Cross-compilation for Android from a Windows host isn't officially supported.

If these constraints are limiting for your use case, please file an issue or reach out on Discord. This workstream is driven by community requests and we will be happy to consider feature and pull requests.

Task List

In addition to the above, there have been a number of community contributions that have laid the groundwork. The following list is not exhaustive, but covers a large chunk of the Windows contributions:

cc @larryliu0820 @jathu

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.