Jules, Google’s AI coding agent, was released from beta on Wednesday, little over two months after its May public peek. Jules is an asynchronous, agent-based coding tool powered by Gemini 2.5 Pro. It interacts with GitHub, clones codebases into Google Cloud virtual machines, and utilizes AI to modify or update code while developers work on other activities.
In December, Jules was launched by Google as a Google Labs project while offering beta testers a chance to use it during the I/O developer conference.Director of product at Google Labs, Kathy Korevec, said that after the tool experienced hundreds of UI as well as quality enhancements during its beta phase, a decision was made to remove it from beta. This decision was motivated by the increased stability of the tool.
She said:
“The trajectory of where we’re going gives us a lot of confidence that Jules is around and going to be around for the long haul”.
Google implemented tiered price tiers for Jules with the broader distribution. The initial “introductory access” free plan was limited to 15 separate daily tasks and three concurrent ones, which was lower than the 60-task restriction during beta. The Google AI Pro and Ultra plans, which cost $19.99 and $124.99 per month and give users 5× and 20× larger limits, respectively, include Jules’ paid tiers.
According to Korevec, packaging and pricing of Jules, were generated using “real usage” facts accumulated over the course of past few months.
Korevec stated that:
“The 60-task cap helped us study how developers use Jules and gave us the information we needed to design the new packaging. The 15/day is designed to give people a sense of whether Jules will work for them on real project tasks.”
In an effort to be transparent about how it trains AI, Google has additionally altered privacy statement of Jules. Data from a public repository could be used for training, but no data is supplied from a private repository, according to Korevec.
Korevec also added that:
“We got a little bit of feedback from users that it [the privacy policy] wasn’t as clear as we thought it was, and so we’re most of it is just responding to that. We didn’t change anything about what we’re doing on the training side, but we changed the language”.
According to Google, thousands of developers worked on tens of thousands of jobs during the beta, producing more than 140,000 publicly available code updates. Based on early feedback, new features were added by the team of Google Lab which includes multimodal input support, integration with GitHub problems, as well as the ability to recycle the previous setups for quicker task execution.
According to Korevec, two main users of Jules are professional developers along with AI enthusiasts. Unlike popular AI coding tools like Cursor, Windsurf, and Lovable, which operate synchronously and require users to view the output after each prompt, Jules operates asynchronously on a virtual machine.
It was further explained by Korevec:
“Jules operates like an extra set of hands… you can basically kick off tasks to it, and then you could close your computer and walk away from it if you want and then come back hours later. Jules would have those tasks done for you, versus if you were doing that with a local agent or using a synchronous agent, you would be bound to that session”.
This week, Jules got a feature called Environment Snapshots that allows it to save dependencies and install scripts as a snapshot for quicker, more reliable task execution. It additionally achieved a deeper communication with GitHub which makes it possible to open pull requests on its own, just like it could open branches.
In accordance with data from market intelligence firm SimilarWeb, which was examined by TechCrunch, Jules has drawn in 2.28 million visitors worldwide since debuting into public beta. 45% of those views are coming from mobile devices. Vietnam and the United States were the next two largest traffic markets, after India.
Information about top regions along with user base of Jules was retained by Google. During the beta, Korevec told TechCrunch, the team noticed that a lot of people were using Jules from regular vibe coding tools to either extend the vibe-coded project to make it more production-ready or correct issues that might have been implemented.
Jules previously insisted that users have an existing codebase. On the other hand, Google quickly noticed that an extensive number of potential users might be happier trying it out without an existing codebase, similar to those exploring other AI technologies. As reported by Korevec, the organization immediately attempted to make it easier for Jules to continue working even with nothing in her repository. As a result, its use and breadth increased.
Additionally, the team of Google Labs discovered that more individuals were utilizing their mobile devices for accessing Jules. In the words of Korevec, people were using the web app of tool even though it did not have an exclusive mobile app.
She acknowledged that:
“Since it’s a big use case that we’re seeing emerging, we’re absolutely exploring what the features are that people need on mobile a lot more”.
According to Korevec, Google already employs Jules to assist with the development of several internal projects in addition to beta testers, and the business is currently under “big push” to use the tool on “a lot more projects.”
