Gemini 4 Argon Review: Features, Benchmarks & Pricing

Gemini 4 Argon is Google's new frontier model and the first of its Gemini 4 generation, announced on September 30, 2026. Google built it for long, complex work in software engineering, finance, legal research, and cybersecurity defense, and gave it an industry-leading 1-million-token output limit. For now, though, almost nobody can use it: access is limited to a vetted group of cyber defenders, while Gemini 3.5 Flash and other Gemini models remain the practical options today.

This Gemini 4 Argon review covers what the model is, what Google's benchmark results show, what it will cost, when it might reach more people, and where its limits sit. It also covers how to use Google's models right now: Gemini 4 Argon is not integrated into Chat Smith yet, but several other Gemini models already are.

Gemini 4 Argon is Google DeepMind's most capable model to date, designed to sustain deep reasoning across long, multi-step workflows. Google describes it as frontier-level at coding, knowledge work, cybersecurity defense, and creative writing, and says thousands of its own employees already use it for debugging, research, and large codebase migrations.

Its headline spec is output length. Argon can generate up to 1 million tokens in a single response, up from the previous 64,000-token limit, which gives it room to think through hard problems and produce very large outputs in one go. Google has not yet published other details, such as its context window.

Argon sits above the Flash models that make up most of Google's current lineup, such as the one covered in our Gemini 3.6 Flash review. If you want to see how Google's models compare with Anthropic's more broadly, our Gemini vs Claude comparison is a good starting point.

Four areas stand out in Google's announcement.

Argon sets a new state of the art on DeepSWE v1.1 with 77.9%, a benchmark of real-world, long-horizon software engineering. Inside Google, Argon agents are migrating C and C++ codebases to Rust, reaching more than 800,000 lines for one kernel project, and one rewritten video decoder now runs 2.7 times faster than the previous Rust port with identical output.

Google says Argon leads the Vals Index, which weights finance, coding, legal, and tax work by their share of U.S. GDP, and performs strongly on Vals Finance Agent v2 and Harvey's Legal Agent Benchmark. On Zapier's AutomationBench, which tests end-to-end business workflows, it ranks first with 51.3%. For planning tasks such as drafting with an AI business plan generator, these are the gains most professionals would notice.

Argon is also strong when work involves visuals: analyzing professional charts, acting on a series of documents, or finding details in long videos. On LVBench, a long-video understanding benchmark, it scores a state-of-the-art 91.7% — the kind of capability behind tools that summarize YouTube videos in seconds.

Google trained Argon to find, validate, and patch critical software vulnerabilities on its own. It ties for first place on CWE-bench v1 with 68%, and security firm Wiz used it to uncover a critical flaw exposing sensitive personal data in healthcare software used by hospitals worldwide — a risk earlier frontier models had missed. Trusted defenders receive a version without cyber guardrails.

The table below collects the key figures from Google's announcement. Benchmark results are Google's own and have not yet been widely verified by independent testers.

The numbers point to a model built for depth: long coding sessions, multi-step business tasks, and security work. Google also reports leading results on the Vals Index, Vals Finance Agent v2, and Harvey's Legal Agent Benchmark.

Google says Argon will launch at an introductory price of $2 per million input tokens and $10 per million output tokens, with cached input discounted by 95%. After the introductory period, the price rises to $4 and $20. These rates apply to API access, not to Gemini app subscriptions.

• Now: rolling out to trusted cyber defenders through Google DeepMind's Fairwind Program, while Google takes part in the U.S. government's voluntary pre-release model access process.
• Next: paid Gemini API customers and Google AI Ultra subscribers, with no date announced.
• Later: broader availability for developers, enterprises, and consumers once Google finishes strengthening its safeguards.
• Chat Smith: Gemini 4 Argon is not integrated yet. You can use Gemini 3.5 Flash and other Gemini models today, with higher usage limits on Chat Smith Pro at $9.99 per month or $39.99 per year.

The biggest limitation is access. There is no public release date, Google's subscription plans do not yet list Argon, and even the next wave of access is limited to paid API customers and the top Google AI Ultra plan.

The benchmark results are also mostly Google's own. Press coverage noted that Argon still trailed rivals on two of the four coding benchmarks Google included, and independent testing at scale will only be possible once more people can use it.

Safety controls will shape the experience too. Argon is designed to refuse requests that could help with cyberattacks or chemical, biological, radiological, or nuclear weapons, and Google monitors its chain of thought and actions so it can stop execution when the model goes beyond what the user intended. The 1M-token output limit is also an API capability — it does not mean every chat reply will be that long.

Gemini 4 Argon is not integrated into Chat Smith yet, but several Gemini models already are. Pick one from the model list and start chatting on Android, iOS, or in the browser, with no Google AI subscription needed.

For everyday work, Gemini 3.5 Flash is the fastest way in: quick answers, writing, research, and summaries. For tasks that need more careful reasoning, such as long documents or complex analysis, Gemini 2.5 Pro gives you a deeper option. For images, Google's Nano Banana Pro handles image generation and editing in the same app.

Because Chat Smith also includes models from OpenAI, Anthropic, xAI, and DeepSeek, you can send the same prompt to a Gemini model and its rivals and keep the answer that holds up. Chat Smith Pro adds higher usage limits and more premium models for $9.99 per month or $39.99 per year.

Gemini 4 Argon is Google's most ambitious model yet: state-of-the-art results in long-horizon coding, business automation, and long-video understanding, leading defensive security skills, a 1M-token output limit, and an aggressive introductory price. Its main drawback is simple — almost no one can use it yet, and Google has given no date for broader access.

Until Argon reaches more people, Google's current Gemini models remain the practical choice. You can try them now in Chat Smith, an AI chatbot that puts Gemini next to the latest models from OpenAI, Anthropic, xAI, and DeepSeek.