Free to read. Sign up to save tools and get alerts when they change. Plus 1,000+ more AI tool profiles.

Sign up free
6 min read·Updated November 11, 2026

Gemini 4 Argon is Google DeepMind's frontier model and its first new flagship since Gemini 3.1 Pro. Announced September 30, 2026 for vetted cyber defenders only, it became generally available on October 13 through the Gemini API and to Google AI Pro and Ultra subscribers in the Gemini app and AI Mode, with agentic coding in Antigravity limited to Ultra. Free users do not get it. Every benchmark on this page is Google's own.

Share

Listen to this overview

Free preview · first 0:30
0:00 / 0:30

Unlock audio and more

Audio streaming, downloadable PDFs and certificates come with Plus and Pro.

Learning Objectives

  • Understand what Gemini 4 Argon is and why Google released it to cyber defenders first
  • Read Google's benchmark claims for what they are: vendor-reported and not yet independently reproduced
  • Know which plans include Argon, what the API costs, and when the introductory price ends

What Is Gemini 4 Argon?

Gemini 4 Argon is Google DeepMind's next frontier model, announced on September 30, 2026 by Koray Kavukcuoglu, Google's chief AI architect. It is the first new flagship since Gemini 3.1 Pro in February. Google had promised a Gemini 3.5 Pro for the summer and shipped a run of Flash models instead; Argon jumps the version number to 4 rather than closing that gap.

Google positions Argon for long, multi-step professional work: real-world software engineering, enterprise knowledge work such as legal and finance, and cybersecurity defense. The most concrete change is length. Argon's output limit is 1 million tokens, up from 64,000 in earlier Gemini models, so a single run can produce hundreds of thousands of tokens of reasoning and output rather than stopping and being restarted.

⚠️Warning

Argon is a paid model. Since October 13, 2026 it is available to developers through the Gemini API and to Google AI Pro and Ultra subscribers in the Gemini app and AI Mode in Search, but agentic coding in Google Antigravity is limited to Ultra, and free users get nothing. Google is keeping Gemini 3.1 Pro available for now. Vetted defenders in the Fairwind Program still get a separate version without the usual cyber guardrails.

From Gated Preview to General Release

For its first two weeks Argon was available only through Google's Fairwind Program, the access gate Google created in early September 2026 for its Gemini 3.8 Flash Cyber model. On October 13 Google opened it to developers and paying subscribers, the step it had promised without a date, and it went further than it had said it would, adding Pro subscribers in the Gemini app alongside Ultra. Google says the public version carries its cyber safeguards, while Fairwind members, now expanding to hundreds more partners, keep the unguarded one.

Google's new examples are internal. It says Argon in Antigravity raised agentic code submissions by more than 35 percent while cutting rollbacks by nearly a third, and helped the Chrome team separate a 6 million-line legacy folder from the rest of Chromium. Ars Technica adds that Argon found redundant emoji rendering in YouTube's comment system whose removal cut server load by 45 percent. These are Google's own accounts of its own codebases, and none comes with outside verification.

Why Cyber Defenders Go First

Google said releasing capabilities at this level safely required a phased approach, and that it is taking part in the US government's voluntary process for testing models before release. Argon was trained to find, validate and patch critical software vulnerabilities on its own, and Google says it is giving the model to trusted defenders and its own teams without its usual cyber guardrails, so they get the full capability. That is the reason for the gate: the same skill that patches a flaw can find one to exploit.

The early example Google cites comes from Wiz, which is using Argon in its free Scan for Good program for critical public infrastructure. Google says the model found a critical vulnerability exposing sensitive personal information in healthcare software used by hospitals worldwide, one that earlier frontier models had missed. Google did not name the software or publish details, so treat it as a claim rather than a demonstration.

Before the wider release, Google said it was strengthening four kinds of safeguard. It improved refusals for cyber and chemical, biological, radiological and nuclear misuse while trying to keep legitimate research working. It is hardening the model against prompt injection, where it says Argon leads Gray Swan's indirect prompt injection benchmark. It runs a monitor that reads the model's chain of thought and actions and stops a run that goes beyond what the user intended. And it is sealing its testing sandboxes before high-risk training and evaluation begin. Google also urges other labs to keep model reasoning readable, so that monitors like this one keep working.

Benchmarks

Every figure below is reported by Google. None has been independently reproduced, and Google published them without the full evaluation settings.

BenchmarkWhat it measuresGemini 4 Argon
DeepSWE v1.1Long-horizon real-world software engineering77.9% (state of the art, per Google)
Vals IndexFinance, coding, legal and tax work weighted by US economic outputFirst place
AutomationBench (Zapier)End-to-end business task execution51.3% (first place)
CWE-bench v1Fixing security vulnerabilities68% (tied for first)
LVBenchLong video understanding91.7%

Ars Technica reports that the DeepSWE score is ahead of GPT-6 Astra, Claude Fable 5.1 and Claude Opus 5.5. Google also names leading results on Vals Finance Agent v2 and Harvey's Legal Agent Benchmark without giving the numbers.

Google's examples of internal use are more specific than the benchmarks. It says a team of Argon agents analyzed memory profiling across Google's data centers and freed more than 300 tebibytes of memory, that Argon agents are migrating C and C++ code to Rust at scales up to the 800,000-line Fuchsia Zircon kernel, and that on Google's open-source video decoder the result runs 2.7 times faster than the earlier Rust port with identical output. Google adds that these rewrites still go through automated and manual auditing before reaching production.

Pricing & Access

Gemini API (introductory)$2 input / $10 output per 1M tokens
  • Through November 30, 2026
  • Cached input 95 percent off
  • AI Studio and Gemini platform
Gemini API (standard)$4 input / $20 output per 1M tokens
  • From December 1, 2026
  • Same model
  • Public version, with cyber safeguards
Google AI Pro$19.99/month
  • Argon in the Gemini app and AI Mode
  • Coding in the app's Canvas mode
  • No Antigravity access
Google AI UltraSubscription
  • Argon in the Gemini app and AI Mode
  • Agentic coding in Antigravity
  • The only plan with Antigravity access
Fairwind ProgramBy application
  • Vetted cyber defenders
  • Released without cyber guardrails
  • Expanding to hundreds more partners

The introductory API price is low for a frontier model. GPT-6 Astra lists at 10 dollars per million input tokens and 50 dollars per million output. The rate doubles once the introductory period ends on November 30, so budget for the standard price on anything you plan to run past then.

Strengths

  • Output length: a 1 million token output limit lets one run carry a long migration, report or analysis without being broken into pieces
  • Long-horizon coding and professional work: Google reports leading scores on DeepSWE, the Vals Index and AutomationBench
  • Cyber defense by design: trained to find, validate and patch vulnerabilities, with a gated release that puts that skill with defenders first
  • Readable reasoning as a safety tool: Google monitors Argon's chain of thought and argues publicly that other labs should keep theirs readable too
  • Low introductory price: the API rate sits well below the closed frontier models it is compared with, at least through November 30

Limitations & Considerations

  • No free access: free Gemini users do not get Argon, and agentic coding in Antigravity requires the Ultra plan
  • Every benchmark is vendor-reported: no independent lab has reproduced Google's numbers, and the most striking claim, the Wiz hospital-software flaw, came with no details
  • Two versions: Fairwind defenders get the model without cyber guardrails while the public version keeps them, so results reported by security teams may not carry over to what you can call
  • No context-window figure published: Google announced the output limit, not the input limit, so do not assume the two match
  • Pricing will change: the introductory rate doubles to 4 dollars and 20 dollars per million tokens after November 30, 2026
  • Not Google's strongest model: Gemini 4 Carbon, available since November 10, 2026 to API developers and Google AI Ultra subscribers, scores higher on the hardest tasks, though Google says Argon remains the better choice for most work

Gemini 4 Carbon: The Second Gemini 4 Model

On October 28, 2026, two weeks after Argon reached paying users, Google previewed Gemini 4 Carbon, which it calls its strongest model, again for Fairwind cyber defenders first. On November 10 it opened Carbon to paying developers and enterprises at the same introductory API price as Argon, and to Google AI Ultra subscribers in the Gemini app, AI Mode and Antigravity.

Carbon does not replace Argon. Google says Argon remains the right choice for the majority of software engineering and knowledge work, while Carbon spends more tokens on the hardest coding, research and security tasks. Ars Technica found Carbon's lead over Argon slim on every test Google published at the preview, and the research firm SemiAnalysis's GDPx benchmark put Carbon in the same range as Claude Opus 5.5 and GPT-6 Astra while rating Opus 5.5 best overall. Google AI Pro subscribers stay on Argon, so for them Argon is still the top Gemini model available.

  • Gemini 4 Carbon — the second Gemini 4 model, for the hardest coding, research and security work
  • Gemini 3.1 Pro — the previous Pro-tier flagship, which Google is keeping available for now
  • Gemini 4 Hydrogen — the fast, lower-cost Gemini 4 model that replaced 3.8 Flash on November 12, 2026
  • Gemini — Google's consumer app, where Argon reaches Google AI Pro and Ultra subscribers
  • GPT-6 Astra — OpenAI's flagship, generally available since September 3, 2026, and the model Google benchmarks Argon against
  • Wiz Cloud Security — the security company using Argon in its Scan for Good program

Lineage

  • Gemini 4 Carbon (previewed October 28, 2026; available November 10) — Google's strongest model, for the hardest tasks
  • Gemini 4 Argon (announced September 30, 2026; generally available October 13) — frontier model; the page above
  • Gemini 4 Hydrogen (November 12, 2026) — the fast, lower-cost model, which Google says edges Argon by one point on the Artificial Analysis Intelligence Index while recommending Argon for long, multi-step work
  • Gemini 3.8 Flash (September 2, 2026) — the previous fast model, with a Fairwind-gated Cyber variant that remains current
  • Gemini 3.5 Pro (announced May 2026) — promised for the summer and never released
  • Gemini 3.1 Pro (February 2026) — the last Pro-tier model to ship

Key Takeaways

  • Gemini 4 Argon is Google DeepMind's new frontier model and its first new flagship since Gemini 3.1 Pro in February 2026, jumping past a Gemini 3.5 Pro that never shipped
  • Argon launched to vetted security teams only and became generally available on October 13, 2026, through the API and to Google AI Pro and Ultra subscribers; free users do not get it
  • Fairwind defenders still get a version without the usual cyber guardrails, while the public version keeps them
  • Google reports 77.9 percent on DeepSWE and first place on the Vals Index and AutomationBench, but every number is vendor-reported and none has been reproduced
  • The output limit rises to 1 million tokens from 64,000, the change most likely to matter for long migrations and reports
  • The API costs 2 dollars per million input tokens and 10 dollars per million output through November 30, 2026, then 4 dollars and 20 dollars
  • Gemini 4 Carbon, Google's strongest model, opened to developers and Ultra subscribers on November 10, 2026, but Google says Argon remains the better choice for most work, and Pro subscribers keep Argon

Keep track of the tools you’re evaluating

  • The AI Hub on a phone: a 12-day AI Skill Streak and an expanded Content updates alert listing the saved items that changed.
  • Recommended for you on a phone: nine personalised suggestions labelled Trending in AI news, On your saved list, and Popular.
  • My AI Tools on a phone: saved tools including GitHub Copilot and OpenAI Codex, each with an Updated badge.

Swipe for Recommended for you and My AI Tools

Your AI Hub — sample data.

📰Gemini 4 Argon in the News

Showing the 5 stories where Gemini 4 Argon is tagged in Top AI Stories.

Other tools in Foundation Models & Open Source (12 of 84)

Show 7 more →

Other tools from Google

Show 21 more →
🧭Recommended for you

Optional detours — these connect to what you just read, and your next lesson will be waiting.