Google’s AI Coding Agent Jules Now Available to the Public: Revolutionize Your Code!

Months after its debut in public beta, Google’s AI agent for software development, Jules, has now been made officially available to the general public. Powered by Gemini 2.5 Pro, this tool has moved beyond experimental stages to become a fully integrated part of Google’s AI ecosystem, featuring structured pricing, expanded capabilities, and extensive user feedback.

Jules: A Standout Asynchronous Coding Agent

Jules is designed to operate independently, performing development tasks on-demand in a cloud-based virtual environment, without the need for continuous developer presence. Google highlights this asynchronous approach as a key differentiator from traditional AI agents such as Lovable, Cursor, or Windsurf.

“Jules acts like an extra pair of hands… You can assign it a task, shut your computer, and return hours later to find it completed,” says Kathy Korevec, a product director at Google Labs, in an interview with TechCrunch.

During its public beta phase, Google reports that thousands of developers executed tens of thousands of tasks, resulting in over 140,000 code enhancements. This usage allowed Google’s teams to fix hundreds of bugs, enhance the user interface, and introduce anticipated features such as the reuse of previous configurations, multimodal support, and improved integration with GitHub and native GitHub Issues.

Public Launch Supported by Stability and Real-World Use

The testing period also revealed new user behaviors. Notably, many users approached Jules without an existing codebase, prompting Google to adapt Jules to also function with an empty repository. Another significant development is the increasing consideration of mobile usage, although there is not yet a dedicated app for the tool.

According to Kathy Korevec, it’s the continuous improvement in stability that justified the move out of beta: “The path we’re on gives us a lot of confidence in Jules’ longevity,” she states. The product team also claims hundreds of updates related to the interface and overall quality of the user experience.

Google has also updated Jules’ privacy policy. While no substantive changes were made, the wording has been revised. Kathy Korevec explains: “We received feedback that it wasn’t as clear as we thought. Most of the adjustments address this.” Specifically, public repositories may be used to train models, but private repositories are not used, and their data is not sent.

New Structured Pricing Plan

With the general availability, Google introduces a clear pricing structure for Jules. The “discovery” access is free but now limited to 15 tasks per day (and 3 simultaneous), compared to 60 tasks per day during the beta. Two paid plans are available for more intensive use:

• Google AI Pro ($19.99/month): 100 tasks per day, 15 simultaneous tasks,
• Google AI Ultra ($124.99/month): 300 tasks per day, 60 simultaneous tasks, designed for large-scale or multi-agent workflows.

These subscriptions are accessible via the platform jules.google. Google notes that these limits were adjusted based on real data collected during the beta: “The limit of 60 tasks allowed us to study how developers use Jules and gave us the information needed to design this new offer,” emphasizes Kathy Korevec.