Google Launches AI Coding Agent Jules Out of Beta With New Pricing Tiers
Google releases its Gemini-powered coding agent Jules to the general public after a successful two-month beta period. The tool now features structured pricing plans and clearer privacy guidelines regarding AI training.
Google officially launches its AI coding agent, Jules, out of beta just two months after its initial public preview at the I/O developer conference. Powered by Gemini 2.5 Pro, this asynchronous tool integrates directly with GitHub and clones codebases into Google Cloud virtual machines to handle code fixes and updates automatically. The company decides to exit beta after applying hundreds of quality and UI updates that significantly improve the tool's overall stability.
With this wider rollout, Google introduces structured pricing tiers that replace the previous beta limits. A free introductory plan offers users 15 individual daily tasks and three concurrent tasks, while paid subscriptions through the Google AI Pro and Ultra plans provide much higher limits for $19.99 and $124.99 a month respectively. A Google product director notes that these specific pricing packages stem directly from real usage data gathered during the initial 60-task beta cap.
Google also updates Jules' privacy policy to provide explicit clarity on how it handles data for AI training. The company confirms that data from public repositories may be used for training, but emphasizes that no data leaves the environment when developers use private repositories. Since its beta debut, thousands of developers use Jules to complete tens of thousands of tasks, resulting in over 140,000 public code improvements and inspiring new features like GitHub issue integration and multimodal input support.