Gemini 4 Argon Release Date: Is It Out, and Who Can Use It?

Gemini 4 Argon was announced on September 30, 2026, and most people still cannot use it. Google's announcement says the model is "rolling out to a set of trusted cyber defenders through our Fairwind Program", and that it will reach developers, enterprises and consumers "as soon as possible", with no date attached^[1]. So the short answer to "is Gemini 4 Argon out?" is: announced, priced, benchmarked, and gated.

That gap is why the searches look the way they do. In the day after launch, the suggestions around the model were almost all about timing and access: "gemini 4 argon release date", "gemini 4 argon is it out", "is gemini 4 argon here now", "gemini argon login", "gemini argon app". This guide answers those questions from Google's own pages, explains who is inside the program today, and lists what you can run while the rollout continues.

TL;DR

Announced: Google published Gemini 4 Argon on September 30, 2026^[1].
Who has it: a set of partners in Google's Fairwind Program, which works with over 650 partners and prioritizes governments, critical infrastructure operators and core technology platforms^[2].
Who is next: developers, enterprises and consumers, "starting with paid API customers and Google AI Ultra subscribers". No date^[1].
Why the wait: Google is strengthening safeguards in four areas first and is taking part in the U.S. government's voluntary pre-release access process^[1].
No download: Gemini 4 Argon is a hosted model. Google has not announced weights, a local version or an open-source release^[1].

What happened on September 30

Google introduced Argon as its new frontier model, built "to sustain deep reasoning across complex, long-horizon workflows" in software engineering, enterprise knowledge work such as legal and finance, and cybersecurity defense^[1]. The launch post came with a full benchmark table, a price, and a list of internal results: Argon agents moving C and C++ codebases to Rust, a libgav1 decoder that runs 2.7x faster than the earlier Rust port, and memory optimizations that freed more than 300 TiB across Google's data centers^[1].

What the post did not include is a public API, a model ID, or a button in the Gemini app. Everything that makes a model usable for a developer is missing on purpose. Google calls the release "a phased approach" and says it will "gather feedback from early testers as we iterate on guardrails"^[1].

So when people search for the Argon release date, they are really asking two separate questions. The announcement date is settled: September 30, 2026. The availability date, the one that matters for building anything, is still open.

Who can use Gemini 4 Argon today

Only organizations inside the Fairwind Program. Google describes Fairwind as a program that "gives trusted partners a critical head start against cyber threats", and says it currently works with over 650 partners globally^[2]. Not every partner gets the model: the program page says "a set of Fairwind Program partners get exclusive access to Gemini 4 Argon", which they can use inside CodeMender, Google's code security agent^[2].

Google lists three groups it prioritizes^[2]:

Group	What Google says they use it for
Governments and national cyber authorities	Defending public-sector networks and citizen services
Critical infrastructure operators	Protecting healthcare, telecommunications, energy and financial networks
Core technology platforms	Securing software foundations used by millions of downstream users

The access rules are strict. Participating organizations may grant Argon access only to internal cybersecurity, incident response or penetration testing teams, must track employee access, need phishing-resistant MFA, and may not share, redistribute or sell access^[2]. Google also runs background checks on applicants^[2].

One detail explains why this group goes first. For trusted defenders and Google's own teams, Google says it is "releasing Argon without cyber guardrails so they can leverage its full frontier-level cybersecurity defense capabilities"^[1]. The version that reaches everyone else will carry safeguards that the Fairwind version does not.

If your organization fits one of the three groups, the program page has an application form. Google says it will respond to eligible partners "as soon as we can"^[2]. For everyone else, there is no waitlist to join.

When developers get Gemini 4 Argon

Google gives an order, not a date. The announcement closes by saying feedback from the first cohort will help strengthen its systems "before we release to developers, enterprises, and consumers, starting with paid API customers and Google AI Ultra subscribers"^[1].

Read literally, that gives three things:

API access comes before free access. "Paid API customers" means the first developer wave is billed per token, not a free tier.
Consumers start at the top plan. Google AI Ultra subscribers are named; lower plans are not.
The price is already set. Google published an introductory rate and the standard rate that follows it. Our Gemini 4 Argon pricing breakdown covers the numbers.

Anything more precise than that, such as "October" or "within weeks", is a guess. I could not find a date in Google's announcement, the Fairwind page or the DeepMind evaluation page^[1]^[2]^[3].

Why Google is holding Gemini 4 Argon back

Google names four areas it wants to strengthen before broad availability^[1]:

Misuse. The model is designed to refuse cyber and CBRN attack requests while keeping legitimate dual-use research possible. Google says it is improving how it monitors the model's internal activations to spot misuse.
Prompt injection. Google calls Argon its most resilient model yet against indirect prompt injection and says it leads Gray Swan's Indirect Prompt Injection benchmark.
Misalignment. Monitors watch the chain of thought and actions and can stop execution when the model goes beyond what the user intended.
Hardened systems. Sandboxed environments are isolated and sealed before high-risk training or evaluation.

Google also says it is "actively engaged in the U.S. government's voluntary process for pre-release model access while we gradually expand access"^[1]. That is a step outside Google's own control, which is one more reason not to expect a firm date soon.

The strongest reason is in the benchmark table. Argon ties for first on CWE-bench v1 at 68.0%, and Google reports clear gains over its 3.8 Flash Cyber model in vulnerability discovery, including on an internal benchmark spanning 20 programming languages^[1]. A model that good at finding flaws is also good at finding them for the wrong people. The phased release is the price of that capability.

Not for the public. "gemini argon login" and "gemini argon app" are among the most common searches around the model, so the short answers:

App: Google says consumer access will start with Google AI Ultra subscribers, but it has not said when Argon appears in the Gemini app^[1].
Login: there is no public sign-in for Argon. Fairwind partners reach it through Google's managed program^[2].
Download: Argon is a hosted model, and Google has announced no weights or local version^[1]. A page offering an Argon download or APK is not from Google.

What to run while you wait

If you came here to build something, the useful move is to set up the workload on a model you can call today and switch later.

Google's own table compares Argon against three models: GPT-6 Astra, Claude Fable 5.1 and Claude Opus 5.5^[1]. All three are generally available. Claude Opus 5.5 is the closest stand-in on several of the tests developers care about: 74.2% on DeepSWE v1.1 against Argon's 77.9%, and it leads Argon on Terminal-bench 4.0, 66.4% to 57.4%^[1]. Our three-way benchmark comparison goes test by test.

On reAPI, Claude Opus 5.5 and Gemini 3.6 Flash are live today through one chat completions endpoint and one API key. When Argon reaches developers, testing it should mostly be a change to the model name. We track the rollout on the Gemini 4 Argon model page.

FAQ

Is Gemini 4 Argon out?

Gemini 4 Argon is announced but not generally available. Since September 30, 2026, Google has been rolling it out only to a set of trusted cyber defenders in the Fairwind Program^[1].

What is the Gemini 4 Argon release date?

Google announced the model on September 30, 2026. It has not given a date for developer, enterprise or consumer access^[1].

Who gets Gemini 4 Argon first?

A set of Fairwind Program partners. Google prioritizes governments and national cyber authorities, critical infrastructure operators and core technology platforms^[2]. After them, paid API customers and Google AI Ultra subscribers come first^[1].

Can I apply for Gemini 4 Argon access?

Organizations can apply to the Fairwind Program through the form on Google's program page. Google vets applicants and limits access inside each organization to security teams^[2].

Will Gemini 4 Argon be free?

Google has announced only paid pricing: an introductory per-token rate and a higher standard rate afterward^[1]. No free tier has been announced.

Is Gemini 4 Argon available in the UK or EU?

Google has not published regional availability for Gemini 4 Argon. Fairwind lists partners globally, but the developer rollout has no announced regions or dates^[1]^[2].

Can I run Gemini 4 Argon locally?

No. Gemini 4 Argon is a hosted model, and Google has not announced downloadable weights^[1].

Planning around the Gemini 4 Argon rollout

Treat September 30 as the day the specification was published, not the day the model shipped. The benchmark table, the 1M-token output limit and the price are all public, so you can plan a migration, a budget and an evaluation set now. What you cannot plan is the day the API opens, because Google has tied it to safeguards work and a government review that it does not fully control.

The practical setup is a workload that already runs on a generally available model, a test set that reflects your real tasks, and a switch that is one line of configuration. When Gemini 4 Argon reaches paid API customers, you can measure it against what you already have the same day, instead of starting from zero.

References

Google. Gemini 4 Argon: our next era of frontier intelligence. Retrieved October 2026 from blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-4-argon
Google DeepMind. Fairwind Program. Retrieved October 2026 from deepmind.google/fairwind-program
Google DeepMind. Gemini 4 Argon model evaluation: approach, methodology and results. Retrieved October 2026 from deepmind.google/models/evals-methodology/gemini-4-argon