Gemini 4 Argon: Price, Access and How It Compares

Google announced Gemini 4 Argon on 30 September and then held it back. The first users were a hand-picked group of cyber defenders, as our roundup of that week reported. On 13 October the wait ended for everyone else who pays: developers on the Gemini API, and Gemini app subscribers on Google AI Pro and Ultra. This guide covers what Argon is, where you can use it, what it costs, and how it stacks up against Claude and GPT-6.

What Gemini 4 Argon is

Argon is Google's frontier model, the top of the Gemini line. Google's announcement aims it at three jobs: long software-engineering tasks, enterprise knowledge work such as legal and finance, and cybersecurity defence. The cyber claim is why the rollout was staged. In Google's words, "Argon can autonomously find, validate, and patch critical software vulnerabilities," and Google gave a version without cyber guardrails only to trusted defenders and its own teams.

The specification is built for long work. The API model card lists a context window of 1,048,576 tokens, shared between input and output. Responses default to 65,536 output tokens, and a mode Google calls long decode stretches that to the full million, which Google calls "an industry-leading" output limit. Argon reads text, images, video, audio and PDFs, writes text only, and supports three thinking levels, search grounding, code execution, function calling and, in preview, computer use.

Who can use Gemini 4 Argon now

Where	Who gets Argon	Cost
Gemini app	Google AI Pro and Ultra subscribers	AI Pro ₹1,950 a month in India
AI Mode in Google Search	AI Pro and Ultra, per VentureBeat	Included in the plan
Gemini API, AI Studio, Android Studio	Paid API tier only, no free tier	$2 / $10 per 1M tokens until 30 Nov
Google Antigravity	AI Ultra subscribers, per Google's developer post	AI Ultra from ₹6,500 a month in India
Gemini Enterprise	Eligible business subscriptions	Per Google Cloud contract

In the app, Google says Argon gives more direct answers with less hedging, reports progress better on long tasks, and adds a switch that turns a chat into a task Gemini keeps working on while you are away. It also connects to more than 40 apps, including Instacart, OpenTable, Adobe and Shopify.

Gemini 4 Argon price in the API

Google's pricing page lists an introductory rate of $2 per million input tokens and $10 per million output tokens, thinking tokens included, until 30 November 2026. From 1 December it rises to $4 and $20. Cached input costs $0.10 per million now and $0.20 later, though Argon supports only automatic caching, not explicit cache control. There is no free tier.

Model	Input per 1M tokens	Output per 1M tokens
Gemini 4 Argon, until 30 Nov	$2	$10
Gemini 4 Argon, from 1 Dec	$4	$20
GPT-6.1 Sol	$2	$10
Claude Sonnet 5.5	$2	$10
Claude Opus 5.5	$4	$20
GPT-6 Astra	$10	$50

Rival prices are as reported by VentureBeat on 13 October. The comparison tells you what Google is doing: for six weeks Argon sells at mid-tier prices, and from December it sits level with Claude Opus 5.5, still a fifth of GPT-6 Astra.

How good is Gemini 4 Argon? Google's numbers and everyone else's

Google's own results are strong where it chose to test. It reports 77.9% on DeepSWE v1.1, which it calls a new state of the art for long software tasks, first place on Zapier's AutomationBench at 51.3%, 91.7% on the LVBench long-video test, and a tie for first on CWE-bench v1 vulnerability repair at 68%. It also says more than 100,000 Google employees use Argon each month through an internal version of Antigravity.

Independent scoreboards, as reported by VentureBeat, are split. Vals AI ranks Argon first on its Vals Index at 68.90%, narrowly ahead of Claude Sonnet 5.5 and Claude Opus 5.5. Artificial Analysis gives it 53 on its Intelligence Index, level with GPT-6 Astra and behind Claude Opus 5.5 at 58. And in Google's own comparison table, Claude Opus 5.5 leads on Terminal-Bench 4.0, 66.4% to Argon's 57.4%. The reviewer Sam Witteveen put it plainly to VentureBeat: "This is a really, really, really good model. But is it the best model out there? I actually don't think it is."

The fair reading is that Argon is a top-three model with a price advantage until December. For agentic coding in the terminal, Claude still leads on the benchmark closest to that work. For long video and very long outputs, Argon is ahead.

Gemini 4 Argon in India: AI Pro and the Jio offer

Google's India subscriptions page lists Argon under Google AI Pro at ₹1,950 a month and under AI Ultra, which starts at ₹6,500. The free plan lists only Google's Flash-Lite model. The ₹399 Google AI Plus plan carries a footnote mentioning Argon, but Google's launch post names only Pro and Ultra, so Plus users should check the model picker rather than assume.

There is a cheaper route for many Indian users. Jio's Google Gemini offer gives 18 months of Google AI Pro free to Jio users over 18 on an unlimited 5G plan of ₹349 or more, claimed once per number through the MyJio app. The Jio page does not mention Argon by name, but since AI Pro includes Argon, anyone with an active claim should see it in the Gemini app. Jio calls the offer limited-time without giving an end date.

Should you switch to Gemini 4 Argon?

• Developers on a budget: test it now. Until 30 November it costs the same as GPT-6.1 Sol and Claude Sonnet 5.5, with a larger output limit than either.
• Teams doing terminal-heavy agentic coding: keep Claude in the comparison; the Terminal-Bench gap is real. Our Claude vs Gemini comparison covers the day-to-day differences.
• Gemini app users on Pro: you already have it; try the tasks switch on something you would otherwise check back on.
• ChatGPT subscribers: our ChatGPT vs Gemini comparison and guide to whether GPT-6 is free show what each plan gets.

FAQ

Is Gemini 4 Argon free?

No. The free Gemini plan lists only Flash-Lite, and the Gemini API has no free tier for Argon. In India the cheapest confirmed route is Google AI Pro at ₹1,950 a month, or free through Jio's 18-month AI Pro offer if you qualify.

How much does the Gemini 4 Argon API cost?

$2 per million input tokens and $10 per million output tokens until 30 November 2026, then $4 and $20 from 1 December. Cached input is $0.10 per million, rising to $0.20.

Is Gemini 4 Argon better than Claude Opus 5.5?

It depends on the test. Argon leads the Vals Index and Google's long-video and long-task benchmarks; Claude Opus 5.5 leads Artificial Analysis's Intelligence Index and Terminal-Bench 4.0. Until December, Argon costs half as much per token.

What is the Gemini 4 Argon context window?

1,048,576 tokens, shared between input and output. Output defaults to 65,536 tokens and can extend to the full window with long decode.

When was Gemini 4 Argon released?

Google announced it on 30 September 2026 and gave it first to trusted cyber defenders. Broad availability in the Gemini API and the Gemini app followed on 13 October 2026.