Docker Desktop 4.42 adds WSL2 GPU support and a faster AI model runner

Docker has released Desktop 4.42, adding native GPU passthrough for developers running Windows Subsystem for Linux 2. The update is designed to make container workflows smoother for users who need hardware acceleration inside WSL2 environments, a common setup for local development on Windows machines.

The new version also refreshes Docker Model Runner, the company’s tool for running AI models locally. Docker says the changes aim to improve the experience for developers testing and deploying large language models on their own machines, without relying entirely on cloud services.

Alongside the AI updates, Docker says it has improved Compose Watch performance, which should help developers see faster feedback when files change during active development. The release is part of Docker’s continued push to blend container tooling with local AI workflows and day-to-day developer productivity.

Docker Desktop 4.42 is now available through the company’s official release channels.