Google reveals Gemini 4 Argon, but you're not allowed to use it

Credit: Google Published Sep 30, 2026, 5:17 PM EDT Simon is a Computer Science BSc graduate who has been writing about technology since 2014, and using Windows machines since 3.1. After working for an indie game studio and acting as the family's go-to technician for all computer issues, he found his passion for writing and decided to use his skill set to write about all things tech. Since beginning his writing career, he has written for many different publications such as WorldStart, Listverse, and MakeTechEasier. However, after finding his home at MakeUseOf in February 2019, he would eventually move on to its sister site, XDA, to bring the latest and greatest in Windows, Linux, and DIY electronics. Sign in to your XDA account We're getting to the point where AI companies can no longer hand out LLMs to the general public. A year or two ago, we saw them push out new models all the time and watched users put them to work; these days, though, the models can be so powerful that they need extensive testing and guardrails before they're given to the public. We saw it with Claude Mythos, which is so mighty that it was only given out to select entities. We then eventually got our hands on the model with variants like Claude Fable and Opus. Now, it seems Gemini has hit the 'Mythos wall', as Gemini 4 Argon is here, and you are absolutely not allowed to touch it. Not yet, anyway. Google Gemini 4 Argon is the most powerful version of the LLM yet It's not quite ready for the red carpet, though Credit: Google As announced in News from Google, the internet giant has revealed Gemini 4 Argon. You won't find it on your personal Gemini dashboard, because, while it's very powerful and internal Google employees are already using it for day-to-day tasks, it's a little too strong for public consumption. See, we're getting to the point where AI models pose a serious threat to the internet. We've already seen several AI models, including Gemini, breaking out of their confines and hacking into other companies' servers to achieve their goal. Handing those same models into the hands of potential criminals is asking for trouble. As such, Google is working to ensure nobody can misuse Argon: Defending against misuse: To prevent bad actors from using Argon for cyber or chemical, biological, radiological, and nuclear (CBRN) attacks, the model is designed to refuse harmful requests while preserving legitimate, dual-use scientific research, as per our Frontier Safety Framework. We are strengthening the robustness of our safeguards for this launch, including improving our techniques to monitor the model’s internal activations to spot misuse. Right now, the model is being handed to trusted defenders via the Fairwind Program, but once the company is confident that nobody can abuse Argon's capabilities, it wants to "release to developers, enterprises, and consumers, starting with paid API customers and Google AI Ultra subscribers." It'll likely be under a different name, much like Claude Mythos becoming Fable for its public release, but Google hasn't revealed any potential name change yet, if any is even planned.