Google Opens Gemini 4 Argon to Developers and AI Subscribers
Google's first Gemini 4 model, Argon, is now available through the Gemini API, AI Studio and paid Google AI plans after a limited preview focused on cyber defenders.

In this article
Google has made Gemini 4 Argon, the first model in its Gemini 4 family, broadly available. Argon was announced on September 30 with access limited to a small group of cybersecurity defenders; on October 13 Google opened it to developers through the Gemini API and Google AI Studio, to businesses through the Gemini platform and Gemini for business, and to paying Google AI subscribers in the Gemini app and AI Mode in Search, according to Google's announcement.
Google pitches Argon as its new frontier model for long, multi-step work in coding, finance, law, cybersecurity and science. Here is what it does, what it costs and who can use it right now.
What is Gemini 4 Argon?
In a post attributed to Koray Kavukcuoglu, SVP at Google DeepMind and Google's chief AI architect, the company says Argon "works longer and harder to solve highly complex scientific, coding, and agentic problems." It is the first Gemini 4 model to ship, and Google says it is "just the beginning" of the family.
The headline claims include:
- Long tasks: a 1 million-token unified input-output limit through the Gemini API, plus a continuation mechanism for very long outputs.
- Coding: inside Antigravity, Google's agentic coding tool, Google reports more than 35% more agentic code submissions and nearly a third fewer code rollbacks.
- Writing: closer adherence to style instructions, more direct answers and less hedging.
- Spatial and visual work: better spatial reasoning and visual coding, such as 3D games and visualizations rendered in a browser.
- Safety: stronger resistance to jailbreaks and prompt injection. Google cites a 0.7% attack success rate on Gray Swan's indirect prompt injection benchmark.
Google also says Argon ranks first on the Vals Index, which covers finance, coding, legal and tax work, and scored 87% on the ERQA embodied-reasoning benchmark. These are Google's own numbers, so expect independent testing to fill in the picture over the coming weeks.
Gemini 4 Argon pricing and availability
| Who | Where you can use Argon |
|---|---|
| Developers | Gemini API in Google AI Studio (global) |
| Businesses | The Gemini platform and Gemini for business |
| Google AI Ultra subscribers | Antigravity, the Gemini app and AI Mode in Search |
| Google AI Pro subscribers | The Gemini app and AI Mode in Search |
API pricing is introductory until November 30, 2026: $2 per million input tokens and $10 per million output tokens, with cached input tokens 95% cheaper. After that, Google says the public model moves to $4 per million input tokens and $20 per million output tokens.
Google is also inviting up to 10,000 scientists to apply for Antigravity's highest usage tier free for six months, and cybersecurity defenders can apply to its Fairwind Program for access without the usual cyber guardrails.
Why the cyber-first rollout matters
Argon's staged launch stands out. Google's September 30 post said the model would reach "a set of trusted cyber defenders" first, with wider access to follow once Google had gathered feedback and refined its safeguards.
The reasoning is straightforward: a model that is very good at reading and writing code is also good at spotting software flaws. Letting defenders use it first gives security teams a head start on finding and fixing bugs before the same capability is in everyone's hands. Expect other AI labs to face pressure to explain how they handle the same trade-off.
What it means for you
For developers, the introductory pricing window is the cheapest time to benchmark Argon against whatever model you use today, especially for long-context or multi-step agent workloads where the 1 million-token limit matters. If agent-style tools are new to you, our explainer on what AI agents are covers the basics.
For everyday users, Argon in the Gemini app and AI Mode requires a paid Google AI Pro or Ultra plan. If you are deciding which assistant is worth paying for, see our roundup of the best AI tools for productivity.
Bottom line
Gemini 4 Argon is what Google calls its new frontier model and, after a cautious cyber-first preview, it is now available to anyone with an API key or a paid Google AI plan. Google's benchmark claims are strong, but the real test will be how it performs on everyday coding, research and writing work once developers and independent evaluators have had time with it. You can follow more launches like this in our AI news section.
Source: Google (The Keyword)
Add The Tech Wonder as a preferred source on Google



0 Comments
Join the discussion. Comments are moderated and your email is never published.