Google releases AI programming assistant Jules in open beta
Google has announced that its new AI programming assistant, Jules, is now available worldwide in public beta and can work on multiple tasks at once.
Google has announced that its new AI programming assistant, Jules, is now available worldwide in public beta and can work on multiple tasks at once.
Google has announced that its new AI programming assistant, Jules, is now available worldwide in public beta and can work on multiple tasks at once.
Jules, first introduced by Google Labs in December, is positioned as a «trusted automated assistant for programming» that can handle a wide range of time-consuming tasks on behalf of users.
As ZDNET notes, the model is «asynchronous» and can launch and work on tasks without waiting for any of them to complete. The AI then provides a full description of the changes that were made to the user’s code to show the reasoning process.
Jules is based on the Gemini 2.5 Pro large-scale language model, which Google introduced in March as «the most advanced model for complex tasks.» Gemini 2.5 Pro outperformed other leading models, including OpenAI’s o3-mini and Anthropic’s Claude 3.7 Sonnet, in key tests such as math and code editing.
This gives Jules some advanced coding capabilities. The assistant can work directly within a user’s codebase, absorbing the full context of the project for decision-making without the need for a «sandbox» (i.e., a separate and controlled testing environment). It also integrates directly with GitHub, eliminating the need for developers to manually switch between coding platforms.
«We are at a tipping point: agent-based development is moving from prototype to product and is quickly becoming a central element in software development», — says Katie Koravek, Director of Product Management at Google Labs.
Jules is currently available for free in public beta, with some usage restrictions.
Let us recall that before this, GitHub presented a new assistant — an AI agent capable of performing programming tasks independently. The tool not only suggests lines of code, but also independently corrects errors, adds new functionality, and updates documentation.
Previously, Jeff Dean, Google’s chief scientist, said that artificial intelligence could soon reach the level of skills of a junior software engineer. He believes that such an AI should know programming better than at a basic level.



