Google on Wednesday launched his AI coding agent, Jules, El Beta, just over two months after his Public preview debut in May.
Operated by Gemini 2.5 Pro, Jules is an asynchronous, agent-based coding tool that integrates with GITHUB, clone code bases in Google Cloud Virtual Machines, and uses AI to repair or update code while developers focus on other tasks.
Google Initially announced Jules as a Google Labs project in December and did it Available to beta testimonies By a public preview at its I/O developer conference.
Kathy Korevec, a product director at Google Labs, told Techcrunch that the improved stability of the tool pushed the decision to extract it from beta after receiving hundreds of UI and quality updates during its beta phase.
“The trajectory from where we go gives us a lot of confidence that Jules is around and going around the long stop,” she said.
With the wider rolling, Google introduced structured prices for Jules, starting with a “introductory access” free plan hatched at 15 individual daily tasks and three simultaneous, down from the 60-task limit while beta. Jules’ paid levels are part of Google AI Pro plans and Ultra, which are price of $ 19.99 and $ 124.99 a month, and offer subscribers 5 × and 20 × higher boundaries respectively.
Korevec noted that Jules’ packages and prices are based on “real use” insights collected over the past few months.
“The 60-task cap helped us study how developers use Jules and gave us the information we need to draw the new package,” she said. “The 15/day is designed to give people a sense, whether Jules will work for them on actual project tasks.”
Google has also updated Jules’ privacy policy to be more explicit about how it trains AI. If a deposit is public, its data can be used for training, but if it is private, Korevec said no data is sent.
“We have received some feedback from users that it [the privacy policy] It wasn’t as clear as we thought, and that’s why we mostly just respond to that. We didn’t change anything about what we do on the training side, but we changed the language, “Korevec said.
During the beta, Google said thousands of developers handled tens of thousands of tasks, resulting in more than 140,000 coding improvements shared publicly. Initial feedback led the Google Labs team to add new skills, including reuse of previous arrangements for a faster task -execution, integrating with GITHUB problems and supporting multimodal input.
The two main users of Jules are the AI enthusiasts and professional developers, Korevec said.
Running asynchronously in a virtual machine, Jules stands apart from top AI code tools like cursor, Windsurf, and lovable, all of which work synchronously and require users to watch the result after each prompt.
“Jules works as an extra set of hands … you can basically launch 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 for you, against if you did with a local agent or use a synchronous agent, you would be tied to that session,” Korevec explained. Korevec.
This week, Jules received deeper integration with GitHub to Open pull requests automatically – just as it could open branches – and a feature called Environmental cartoons To save dependencies and install scripts as an image for a faster, more consistent task -ecute.
From vibrant encoding to mobile use, Beta trials informed Jules Development
Ever since he entered public beta, Jules has registered 2.28 million visits worldwide, 45% of them from mobile devices, according to market intelligence supplier data SimilarWebrevised by Techcrunch. India was the highest market of traffic, followed by the United States and Vietnam.
Google did not share specifications on Jules’ user base and its main geographies.
Korevec told Techcrunch that during the beta, the team observed that many people used Jules from traditional vibrant code tools to repair errors that may have been implemented or extended the VIBE coded project to make it more ready production.
Originally, Jules demanded that users have an existing codebase. But Google soon realized many possible users – like those who try other AIs – may want to explore it without one. Korevec said the company quickly enabled Jules to work even with an empty deposit. This helped increase its scope and use.
The Google Labs team also noticed a growing number of users accessing Jules with their mobile devices. Although the tool has no dedicated mobile program, Korevec said users have accessed it with its -App website.
“Because it is a great case that we see an appearance, we are absolutely researching what the functions that people need with a lot more,” she noticed.
In addition to beta -witnesses, Korevec has stated that Google is already using Jules to help develop some projects inside, and there is now a “big push” to use the “much more projects” tool at the company.
