Gemini 4 Argon Is Now Available: Price, Plans and Limits
Gemini 4 Argon is generally available: Pro and Ultra in the Gemini app, paid API at $2/$10 per 1M tokens until Nov 30, then $4/$20. Plans, limits, Copilot.
PromptWises is reader-supported. Some links may earn us a commission at no extra cost to you. This never affects our ratings or recommendations.
Gemini 4 Argon became generally available on October 13, 2026, two weeks after Google announced it to a small group of testers. If you pay for Google AI Pro or Ultra, it is rolling out in the Gemini app now. Developers can call it as gemini-4-argon on the paid tier of the Gemini API, at an introductory $2 per million input tokens and $10 per million output tokens that doubles on December 1. Free users and the free API tier do not get it. Sources: Google’s announcement, the Gemini app release notes and the Gemini API changelog.
Who can use Gemini 4 Argon
- Gemini app: rolling out globally to Google AI Pro and Ultra subscribers. Google’s blog adds AI Mode in Search for the same two plans.
- Gemini API: paid tier only, through AI Studio and Android Studio. The pricing page says Argon “is not available on the Free tier or on the Google AI Pro plan”, so a Pro subscription does not cover API calls.
- Google AI Ultra: also gets Argon in Google Antigravity, Google’s agentic coding tool, and in AI Studio.
On September 30 Argon was limited to trusted cyber defenders. That earlier programme, Fairwind, is now expanding to hundreds more partners.
API pricing and limits
Prices are per million tokens on the paid tier, at the time of writing:
| Gemini 4 Argon | Through Nov 30, 2026 | From Dec 1, 2026 |
|---|---|---|
| Input | $2.00 | $4.00 |
| Output (including thinking tokens) | $10.00 | $20.00 |
| Context caching (implicit only) | $0.10 | $0.20 |
| Batch or Flex input / output | $1.00 / $5.00 | $2.00 / $10.00 |
Google’s blog footnote says the introductory period “expires on November 30”. The pricing page is more precise: the lower price applies through November 30 and the higher one starts December 1. Search grounding includes 5,000 free requests a month, shared across Gemini 3 and newer models, then costs $14 per 1,000.
The model page lists a context window of 1,048,576 tokens. That figure is a unified limit for input and output together, so a long answer uses up room your prompt could have had. By default the maximum output is 65,536 tokens. With “long decode”, output can use up to the full 1M.
Argon in GitHub Copilot
GitHub made Argon generally available in Copilot on the same day for Copilot Pro+, Max, Business and Enterprise. You pick it in the model picker in VS Code, Visual Studio, JetBrains IDEs, Xcode, Eclipse, Copilot CLI, the coding agent and the Copilot app. It is billed at provider list price under usage-based billing, and the rollout is gradual. In Business and Enterprise, new models are on by default unless an admin turned that off, so admins should check the model policy.
Also new in the Gemini app: Customize and tasks
The release notes also add a Customize item to the left menu. It gathers your response preferences, connected apps and skills in one place. Google says you can connect more than 40 apps and services, including Instacart, OpenTable and Adobe. Most are US English only. Customize is rolling out on web, mobile and the Windows and macOS desktop apps for all tiers. Google Antigravity is now built into the Gemini macOS app.
Google’s app announcement also adds a toggle in the chat window that switches from chat to tasks. In tasks mode Gemini keeps working while you are away from your computer. It is offered to Pro and Ultra subscribers, and Google has not published usage limits for it.
Skills: app rollout started, Workspace delayed
Skills, Google’s replacement for Gems, began a gradual rollout in the Gemini app on October 13 for personal and Workspace accounts. Google expects it to finish by mid-November. The rollout of skills inside Workspace apps is “delayed until further notice”, and Google gave no new date, according to the Workspace Updates post. Our Gemini skills guide explains how to set them up.
Benchmarks: Google’s numbers
Google says Argon ranks first on the Vals Index, which covers finance, coding, legal and tax work, and scores 87% on ERQA, an embodied reasoning test. Both figures come from Google, on benchmarks Google chose. Our guide to reading AI model announcements explains why you should try a model on your own tasks before you switch.
What it means for you
If you already pay for Google AI Pro, try Argon on your hardest work this week. For deciding between assistants, our Gemini review, Claude vs Gemini and ChatGPT vs Gemini comparisons now mention Argon, and our ChatGPT vs Claude vs Gemini hub compares all three plans side by side. More launches are in our news section.
Gemini 4 Argon is now also available in Perplexity Computer for Pro and Max subscribers; see our Perplexity Computer update. Our Gemini review covers which Google AI plan is worth paying for.
Frequently asked questions
Can I use Gemini 4 Argon on the free Gemini plan?
No. In the Gemini app it is rolling out to Google AI Pro and Ultra subscribers only, and in the Gemini API it is not available on the free tier. Google AI Plus is not listed either.
How much does the Gemini 4 Argon API cost?
At the time of writing, Google's pricing page lists $2 per 1M input tokens and $10 per 1M output tokens (thinking tokens included) through November 30, 2026, and $4 and $20 starting December 1, 2026. Batch and Flex are half price.
Does a Google AI Pro subscription include Gemini 4 Argon API usage?
No. Google AI Pro gives you Argon in the Gemini app and AI Mode in Search, but the API pricing page says Argon is not available on the Google AI Pro plan. API use needs a paid API billing account.
Which GitHub Copilot plans include Gemini 4 Argon?
Copilot Pro+, Max, Business and Enterprise. It is billed at provider list price under usage-based billing, and Business and Enterprise admins control it through the Copilot model policy.
Is Gemini 4 Argon the best AI model?
Google says it ranks first on the Vals Index and scores 87% on the ERQA embodied reasoning benchmark. Those are Google's own claims on benchmarks it chose, so test it on your own work before switching.
