Google has moved Jules, its asynchronous, agent-based coding assistant, out of beta and into general availability, positioning it as a tool for developers who want to offload routine programming tasks. Powered by the Gemini 2.5 Pro model, Jules is designed to handle a wide range of coding activities, from writing tests and building new features to fixing bugs, generating audio changelogs, and updating dependencies.
Unlike traditional real-time code assistants, Jules operates asynchronously. It integrates directly with a developer’s existing repositories, clones the codebase into a secure Google Cloud virtual machine, and works in the background. Once tasks are complete, Jules provides a plan, its reasoning, and a diff of the changes, allowing developers to review and approve before merging. Google says the service is private by default, does not train on users’ private code, and keeps all data within its isolated execution environment.
The launch follows a beta period in which thousands of developers completed tens of thousands of tasks using Jules, resulting in over 140,000 code improvements shared publicly. Based on user feedback, Google refined the interface, fixed hundreds of bugs, and introduced new features, including faster task execution by reusing prior setups, GitHub Issues integration, and multimodal support.
With the public launch, Jules now comes in three access levels. The base tier is aimed at trying out the assistant on smaller projects. Google AI Pro is designed for sustained daily work, and Google AI Ultra is built for high-intensity coding environments that require large-scale, multi-agent support.
Not all early testers have been fully satisfied. On Hacker News, one user noted:
I’ve been playing with it, and I’ve been generally not impressed. There are both obvious annoying UI bugs (which should be easy to fix unless they vibe coded the whole thing), and the output of the tool isn’t very good for anything but the simplest problems. If the model was really good, I’d love this, but it’s not.
Some have also voiced concerns over the complexity of Google’s broader AI and Workspace offerings. User Lucasoato commented:
We’ve been trying to understand Google Workspace subscriptions, but it’s a complete mess. I can’t even tell if we have access to Google AI Studio or not. Their tutorials are complete [nonsense], the docs are just plain wrong because they reference things not reflected in the platform.
Jules is available through Google AI Studio, though the exact limits and capabilities depend on the user’s Google AI subscription tier. Google has not announced any changes to simplify the subscription model, but says it will continue iterating on Jules and integrating developer feedback into future updates.
