Google Open-Sources a VS Code Extension That Runs Cloud Workbench Notebooks Without Leaving the Editor
Google's new Apache-2.0 VS Code extension connects a local editor to managed Jupyter Workbench instances on Google Cloud.
Overview
Google has released a Visual Studio Code extension that lets developers run Jupyter notebooks directly on its managed cloud Workbench instances without switching out of their local editor. Announced on the Google Developers Blog, the Workbench Notebooks extension is fully open-sourced, and the company is inviting outside contributions.
The pitch targets a familiar friction point in machine-learning work. According to the Google Developers Blog, “For data scientists and developers, the ideal workflow combines the familiarity of a local IDE with the heavy-lifting capabilities of the cloud.” The extension aims to bridge that gap so that, in the blog’s words, “By eliminating context switching, developers can move from local experimentation to high-performance cloud compute without disruption.”
What We Know
The extension lets developers, per the Google Developers Blog, “Easily connect your local VS Code environment to managed cloud environments” and “Run notebooks directly on Workbench instances without leaving your IDE.” The underlying platform, which the same post describes as having “long been a go-to platform for managed Jupyter environments optimized for data science,” is now branded Gemini Enterprise Agent Platform Workbench.
The project is published on GitHub as colab-enterprise-vscode under the GoogleCloudPlatform organization, licensed Apache-2.0 and written almost entirely in TypeScript. Its repository still describes it as “A Visual Studio Code extension for Vertex AI Workbench” — a reminder that Google renamed Vertex AI to the Gemini Enterprise Agent Platform earlier this year, and the older name persists in some artifacts.
The extension is built on top of the existing Jupyter extension, according to the GitHub repository. Setup routes through the editor’s kernel picker: the repo’s quick start instructs users to “Select Kernel > Google Cloud > Workbench,” then authenticate and pick a Google Cloud project. Once connected, the GitHub repository notes that developers can “interact with other Google Cloud services from within a Workbench instance’s Jupyter notebook.” The repository also states that the extension collects no client-side telemetry.
Distribution runs through two channels. Google directs users to “Download the extension from the VS Code Marketplace today, and contribute to the project on GitHub,” according to the Google Developers Blog, and the GitHub repository notes it can also be installed from Open VSX, the vendor-neutral registry used by VS Code forks. On the Visual Studio Marketplace, the listing is published by “Google Cloud (GoogleCloudTools)” as version 0.1.1 with the description “Connect notebooks to Workbench instances,” and is marked as a free Preview in the Notebooks category.
What We Don’t Know
Google’s post does not specify a license in its own text; the Apache-2.0 designation comes from the GitHub repository. The company has not published a roadmap for moving the extension out of Preview, nor detailed which additional Google Cloud services the in-notebook integration will cover beyond what the current release supports. Adoption is early: the Visual Studio Marketplace lists version 0.1.1 as a Preview release.