Gemini 4 Argon defaults to on for Gemini Enterprise users from 27 October 2026
From 27 October 2026, the Gemini 4 Argon toggle flips to on by default in Gemini Enterprise. Here's what admins need to know.
If your organisation uses Gemini Enterprise, something changed on 27 October 2026 that you need to be aware of, particularly if you are an IT admin. Google’s new Gemini 4 Argon model, which became generally available on 13 October 2026, had its feature toggle set to off by default at launch. From 27 October 2026, that toggle flips to on automatically for all Gemini Enterprise web and mobile app users who had not already enabled it. If you have not reviewed your admin settings, your users now have access to a new frontier model.
What is Gemini 4 Argon?
Gemini 4 Argon (gemini-4-argon) is Google’s most capable model to date, described as being built for “complex coding, agentic, math, and science tasks.” It is the first Gemini 4 model, and notably Google did not use a “Pro” naming convention here. The model features a 1 million token context window, supports text, image, video, audio, and PDF inputs, and has a knowledge cutoff of March 2026. Its architecture is a sparse mixture-of-experts, trained from scratch rather than fine-tuned from an earlier version.
Google first made Argon available to a limited set of partners on 30 September 2026 through its Fairwind Programme, before opening broad general availability on 13 October 2026.
What changed on 27 October 2026
At GA launch, the Gemini 4 Argon toggle in Gemini Enterprise was set to off by default. Admins who wanted to give users access had to manually enable it via the Manage web app features settings.
From 27 October 2026, Google flipped that default. The toggle is now on by default in the global, us, and eu regions. Any Gemini Enterprise organisation that did not pre-emptively enable or disable it before that date will find that Gemini 4 Argon is now available to their users in the web and mobile apps.
To disable the model, admins need to navigate to the Gemini Enterprise feature management settings and switch the Gemini 4 Argon toggle off. The official documentation outlines the exact steps.
What this means for your users
For most Gemini Enterprise users, the practical change is that Gemini 4 Argon now appears as a model option in the Workspace apps they already use. Users are not forced onto it; they can still select which model they use in the chat interface. The model becoming available by default is not the same as it being selected automatically for every conversation.
That said, users may naturally gravitate toward it once they notice it is available, particularly for complex tasks where Argon’s capabilities are well suited.
The billing picture, and where it gets nuanced
This is the point worth reading carefully. For standard Gemini Enterprise app usage, in-app conversations with Gemini 4 Argon are covered by your existing Gemini Enterprise licence and its associated quotas. There is no separate per-token charge for users accessing the model through the Workspace web or mobile apps. The billing model here is subscription-based, not consumption-based.
The per-token pricing is relevant in two other scenarios:
API access. Developers calling gemini-4-argon via the Gemini API are charged at $2.00 per million input tokens and $10.00 per million output tokens through 30 November 2026. From 1 December 2026, those rates double to $4.00 and $20.00 respectively.
AlphaEvolve. From 26 October 2026, Gemini 4 Argon replaced Gemini 3.8 Flash as the default model for new AlphaEvolve experiments. If your team runs AlphaEvolve experiments without specifying a model in the models field of generationSettings, those experiments are now running on Gemini 4 Argon and are billed at Argon token prices. To keep using Gemini 3.8 Flash, explicitly set gemini-3.8-flash in the models field.
AlphaEvolve pricing for Gemini 4 Argon on the Agent Platform is $4.00 per million input tokens and $20.00 per million output or thinking tokens at the model level, with an additional agent surcharge on top of that.
What admins should do right now
A few concrete steps worth working through:
Check your toggle status. Log in to the Google Cloud Console or Workspace Admin Console and review your Gemini Enterprise feature management settings. Confirm whether the Gemini 4 Argon toggle is on or off, and whether that reflects your organisation’s intent.
Review AlphaEvolve usage. If your team uses AlphaEvolve for coding experiments, check whether experiments are specifying a model explicitly. Any experiment relying on the default model is now running on Argon and accruing higher token costs.
Communicate with your users. If you are keeping the toggle on, it is worth letting users know that Argon is now available and giving them a sense of where it is most useful. If you are disabling it, users will benefit from knowing why and what the timeline might look like.
Plan for the December pricing change. If your organisation uses the Gemini API directly, the introductory pricing for Gemini 4 Argon ends on 30 November 2026. The rates double from 1 December 2026, so now is a reasonable time to review API usage and model selection across any internal tools or integrations.
One more thing: single-user quota limits
Separately, on 22 October 2026, Google released a new admin control for Gemini Enterprise Standard, Plus, and Standard Emerging Market editions. Administrators can now configure single-user quota limits on the Usage and Spending page, preventing an individual user from consuming the entirety of an edition’s shared subscription quota pool. This is a useful guardrail, particularly for organisations where usage is concentrated among a small number of power users.
The Gemini Enterprise release notes have the full detail on both of these changes.