Google now allows you to use Gemini 4 Argon, and it wants you to put it to work on your most complex tasks.

Just a few days ago, Google unveiled Argon, its new flagship AI model, although there was a catch. That announcement was simply a staged reveal of a model that was not yet available to the general public: a model available only to researchers and trusted partners that, according to Google, was built for long and complex tasks: coding, intensive office work such as legal and financial tasks, and cybersecurity.

Now, the big news is that Google has released it to (part of) the general public.

Pricing and availability of Gemini 4 Argon

The new Gemini 4 Argon is now available for developers through the API in Google AI Studio, as well as for Google AI Ultra subscribers in Google Antigravity, and for Google AI Ultra and Pro subscribers in the Gemini app and AI Mode in Search. In other words, Argon is already influencing Google Search results, although as noted, only for some users on advanced plans.

According to Google’s documentation, this new model “combines state-of-the-art reasoning, deep visual perception, expanded expert knowledge, and stronger writing capabilities across legal, financial, scientific, and coding domains to support demanding professional workflows. It communicates clearly, keeps you informed every step of the way, and follows style instructions better than any of our previous Gemini models. Argon answers questions more directly and concisely, avoids unnecessary repetition, and leaves clichés out of its responses.”

In a post on X, Robby Stein, Google’s Vice President of Product, explained that “designed to solve your hardest problems, Argon brings big improvements to tackling nuanced, multi-step tasks in AI Mode. It’s better at understanding your intent and following your instructions in back and forth conversations.” You can see how to select it in this video:

Gemini 4 Argon is now rolling out globally in Search!

Built to solve your hardest problems, Argon brings big improvements to tackling nuanced, multi-step tasks in AI Mode. It’s better at understanding your intent and following your instructions in back and forth conversations.

Available starting today for Google AI Pro & Ultra subs, tap the ➕ icon to switch over to Argon and test it out.

— Robby Stein (@rmstein) October 13, 2026

A top-tier model on the global stage

While we wait for the model to become available to all users, the truth is that benchmark platforms such as Artificial Analysis already place it among today’s leading references. In its Intelligence Index, Gemini 4 Argon appears with a score of 53, in sixth place, ahead of the lightweight version of Meta’s new Muse Wonder, and tied with Claude Fable and GPT 6 Astra, two internationally recognized benchmark models. Claude Opus 5.5 leads that ranking, with a score of 58.

The model launches at a starting price of $2 per million input tokens and $10 per million output tokens. That said, Google has already announced that these rates will double starting December 1, 2026, rising to $4 and $20, respectively.

Image: ChatGPT