Google Rolls Out Jules To Rival OpenAI’s Codex

Google is now rolling out the beta version of its coding agent, Jules, which was officially announced last December. 

The agent will fix bugs, build new features, and run and validate changes through unit tests. It is also integrated with GitHub to understand the codebase and works asynchronously. The agent will use Google’s latest Gemini 2.5 Pro model. 

Jules operates within a virtual machine that clones the user’s code, installs necessary dependencies, and alters files. The Jules virtual machine runs Ubuntu Linux and includes developer tools like Node.js, Python, Go, Java, and Rust. Google also mentioned that Jules will not train on private repositories. 

Google is rolling out its own „Codex“ (SWE-Agent), named Jules! pic.twitter.com/2JJpMU586Y https://t.co/Jr5pgBUkpt

— Chubby♨️ (@kimmonismus) May 19, 2025

The tech giant has also published detailed documentation that provides a walkthrough of using the tool. 

Several users across the internet have reported having access to the beta version, which provides five free tasks a day. 

Google’s Jules will now compete with similar capabilities offered by OpenAI. A few days ago, OpenAI launched Codex, a cloud-based software engineering agent. It’s available starting for ChatGPT Pro, Team, and Enterprise users at $200 a month, while it may take a while for the Plus users to get access.

Codex is powered by codex-1, a variant of OpenAI’s o3 model explicitly tuned for software engineering.

Max Weinbach, an analyst at Creative Strategies, said on X, “The Jules coding agent from Google is WAY better than ChatGPT Codex right now. Less lazy, more collaborative, and significantly better quality, it seems.”