Gemini 4 Argon: Proposal Decks, Access, and Alternatives (2026)

Powerdrill Bloom·October 14, 2026

Gemini 4 Argon is Google's new frontier reasoning model. Google made it generally available on October 13, 2026, about two weeks after a limited first release. In Google's own Gemini app demo, Argon turns a folder of project files into a proposal deck for a client.

This guide covers what changed, who can use Argon today, and how to build a proposal deck with it in the Gemini app. It also covers the limits Google lists and the main alternatives. Every detail comes from vendor pages read on October 14, 2026.

What Google released on October 13

Google first announced Gemini 4 Argon on September 30, 2026. That post said the model was "rolling out to a set of trusted cyber defenders through our Fairwind Program." Google said it would widen access after more work on its guardrails.

The wider release came on October 13. The Gemini API release notes list Gemini 4 Argon as "generally available (GA)" on that date. They describe it as "our most capable model built for complex coding, agentic, math, and science tasks." The model ID is gemini-4-argon.

Google's launch post names the places to use it on day one:

Developers and enterprises get the Gemini API and Google AI Studio.
Google AI Ultra subscribers get Google Antigravity, which Google calls its "agent-first development platform."
Google AI Pro and Ultra subscribers get the Gemini app and AI Mode in Search.

The same post sets out the model's focus. It says "Argon excels at deep work in complex, real-world tasks across domains including coding, finance, legal, cybersecurity, and science."

What's new in Gemini 4 Argon

Google's API pages and launch posts list these changes:

One large window. "Input, thinking, and output share one window of 1,048,576 tokens," the API guide says.
Long answers. The default output limit is 65,536 tokens. With long decode, one response can run up to 1M tokens minus the input.
Inputs. The model page lists "Text, Image, Video, Audio, and PDF" as inputs. The output is text.
Thinking levels. Developers can choose low, medium, or high. The API default is medium.
Computer use. The model page marks it as "Supported (Preview)."
Writing style. Google says Argon "answers questions more directly with less hedging" and "skips clichés throughout the responses."

Google also cites results for business work. Its September post says "Argon is the leading model on the Vals Index," which covers finance, coding, legal, and tax work. On Zapier's AutomationBench, the same post reports a score of 51.3%. These are Google's own figures, so test the model on your own files before you switch.

The model card adds one date worth knowing. It says "The knowledge cutoff for Gemini 4 is March 2026." For anything newer, give Argon the source files or let it search the web.

How Argon handles decks and long documents

Google's business example in the Gemini app is a proposal deck. In the Gemini app demo, Gemini 4 Argon starts from "an unstructured folder full of project context." It then "formats it into a clean proposal deck for a client using a custom skill."

The demo continues past the slides. Gemini "drafts and emails the deck to a client." It also builds "an automated workflow for all future incoming client requests with a scheduled task." If skills are new to you, our explainer on Gemini skills covers how they replace Gems.

Documents are the other half of the job. The launch post says Gemini 4 Argon "excels in deep multimodal document comprehension, longitudinal data analysis, and interactive analytical sandboxes." The model card lists "multimodal document analysis and synthesis" among the uses it is "particularly well-suited for."

One customer story shows the scale. At Isomorphic Labs, Google says "Argon is able to quickly synthesize complex PDFs and patents to extract pertinent information." According to the post, that cut "a process that typically takes weeks into hours."

Keep one detail in mind. The API model page lists image generation as "Not supported." In the Gemini app, the tasks feature can create the deck files, and Gemini Apps Help says tasks can use Nano Banana for images. Our guide to Nano Banana 2.1 covers the image model Google released on October 6.

Gemini can also build decks inside Workspace apps such as Gmail and Drive. Our guide to Gemini cross-app decks covers that route.

Who can use Gemini 4 Argon

Access depends on the surface and the plan. This table combines the launch post, Gemini Apps Help, and the API pricing page.

Surface	Who can use Argon	Note from Google's pages
Gemini app chats	Google AI Pro and Ultra	The help center's model table lists Argon for these two plans
Gemini app tasks	Google AI Pro and Ultra	Personal Google Account, age 18 or over
AI Mode in Search	Google AI Pro and Ultra	Named in the launch post
Gemini API and AI Studio	Paid API tier; Ultra subscribers can try AI Studio	API pricing page: not on the Free tier or the Google AI Pro plan
Google Antigravity	Ultra subscribers and enterprises	Named in the launch and developer posts
Gemini for business	Enterprises	Named in the launch post

The two Pro lines cover different surfaces. The Gemini app gives Pro subscribers Argon. The API pricing page adds "Gemini 4 Argon is not available on the Free tier or on the Google AI Pro plan." The developer post likewise names only Ultra subscribers for AI Studio.

The help center's table does not list Argon for Google AI Plus or for accounts without a plan. For work or school accounts, Google says "model and feature availability will depend on your Workspace license." The Google Cloud side is covered in our guide to the Gemini agent for work.

For plan prices, check Google's plans page in your own country. In the API, the pricing page lists per-token rates for Standard, Batch, Flex, and Priority use. Its free tier column reads "Not available" for Argon.

How to make a proposal deck with Gemini 4 Argon

These steps follow Gemini Apps Help for tasks. They assume a personal Google account on a Google AI Pro or Ultra plan.

Start a task. Go to gemini.google.com or open the Gemini app on Mac. Below the text box, click Task.
Pick the model. Open the model menu and choose Argon. Then pick a thinking level, since Gemini Apps Help says it will "default to Low" otherwise.
Add your files. Click Upload & tools, then Upload files, Add from Drive, or Add Notebook. Include the client brief, the pricing sheet, and a past proposal.
Describe the deck. Name the client, the audience, the sections, and the file type. To reuse a skill, type @ and select it; Google's schedules page uses / instead.
Review the files. Open the task's work panel to see which files Gemini read or updated. Ask for changes in the thread, then click Open and Download.
Add a schedule if needed. Ask Gemini to repeat the work every week, or when a matching email arrives.

The help center lists the file types a task can produce. They include .pptx presentations, .docx documents, .xlsx spreadsheets, and PDFs. It adds that "Gemini can also create Google documents, presentations, and spreadsheets."

A sample task prompt shows the pattern: "Use the files in this folder to build a 10-slide proposal deck for Northwind Freight. Include the problem, our approach, timeline, pricing, and next steps. Return it as a .pptx file." The company name is illustrative.

Schedules come in three types. Google lists time-based schedules, Gmail monitors, and topic monitors. In its words, "A schedule acts as the automated trigger for a task, telling Gemini exactly when to step in and execute your instructions."

Before you send the deck, check every price and date on the slides against the source files. The model card notes "general limitations of foundation models, such as hallucinations."

Gemini 4 Argon vs. Gemini Pro in the Gemini app

Pro and Ultra subscribers can choose between Argon and Gemini Pro in chats. The plans page lists access to Gemini 3.1 Pro on both plans. Gemini Apps Help describes the two this way:

	Gemini Pro	Gemini Argon
Google's description	"An advanced reasoning model designed for complex prompts, sophisticated coding, and in-depth problem-solving"	"Top-tier intelligence, dynamic deep reasoning, and the highest performance for complex workflows"
Plans with chat access	Plus, Pro, and Ultra	Pro and Ultra
Listed for tasks	No	Yes

The choice affects your limits. The help center says "More advanced models and higher thinking levels require more of your usage." When you reach the limit, "your chat will continue with Flash-Lite."

A practical split follows from that. Use Gemini 4 Argon for the deck draft and the long documents behind it. Switch to a lighter model for quick edits and short questions.

Limits Google lists

Google's pages spell out several limits for Gemini 4 Argon and for tasks. Each one shapes how you plan a deck workflow.

Accounts. Tasks need a personal account. The help center says "this feature isn’t available if you sign in with a work or school Google Account."
Regions. Tasks are unavailable in the European Economic Area, Nigeria, Switzerland, and the United Kingdom.
Running tasks. "You can have up to 15 tasks running at a given time."
Experimental status. The help center says "Gemini’s ability to handle tasks is experimental." It also warns about prompt injection hidden in documents and websites.
Speed. The model card mentions "occasional slowness or timeout issues." The API guide adds "Hard prompts can take several minutes."
API settings. The API guide says explicit context caching is not supported. It also says to remove temperature, top_p, and top_k from requests.

For a personal account in a supported country, none of these block a proposal deck. They do argue for keeping your numbers in a file you can check, and for reading every slide before it goes out.

Gemini 4 Argon alternatives

Several other models and tools are worth comparing. The table lists what each vendor's pages describe.

Option	What it is	Price on the vendor's page	Where you use it
Gemini 4 Argon	Google's frontier reasoning model	Paid API tier only; rates on Google's pricing page	Gemini app, AI Mode, Gemini API, Antigravity
Claude Opus 5.5	Anthropic's model "For long-running agentic coding and knowledge work"	$4 input and $20 output per 1M tokens	Claude Platform and major clouds
GPT-6 Astra	Listed among OpenAI's flagship models	$10.00 input and $50.00 output per 1M tokens	OpenAI API
GPT-6.1 Sol	Listed among OpenAI's flagship models	$2.00 input and $10.00 output per 1M tokens	OpenAI API
Powerdrill Bloom	An AI workspace that turns files into analysis, charts, and slides	Free plan; Pro $199.00 per year	Web workspace

Claude Opus 5.5. Anthropic's model overview lists a 1M-token context window and 128K tokens of output. Its launch post says "Opus 5.5 communicates more naturally than prior models." It also says the model is "now available on all platforms, including Amazon Web Services, Google Cloud, and Microsoft Azure." Our explainer on Claude Opus 5.5 covers the release.

GPT-6 Astra and GPT-6.1 Sol. OpenAI's API pricing page lists both under its flagship models. The prices in the table are the short-context rates. Our explainer on GPT-6 Astra covers its computer-use features.

Powerdrill Bloom. It starts from your files rather than from a chat prompt. The next section explains where it fits.

Where a file-first workspace fits

A proposal deck depends on the numbers inside it. Those numbers usually live in a pricing sheet, a PDF brief, or last quarter's report.

Powerdrill Bloom starts from those files. Its AI proposal generator page describes turning "a discovery call note and pricing sheet into a branded sales proposal." The outputs it lists include a "Slide partnership pitch deck with mutual value and timeline." The page says you can export a "PDF or PPTX."

For documents you already have, the Word to slides page says the tool is "Perfect for reports, research papers, proposals, and more." The PDF to PPT page describes slides with "summarized insights, clear storylines, and auto-created charts."

The plans spell out the deck side. The Powerdrill Bloom pricing page lists "Create basic slides, docs, sheets, and images" and "1 scheduled task" on the Free plan. The Pro plan, at $199.00 per year, lists "Create slides, Office docs, Excel analysis, and Nano Banana images" and "20 scheduled tasks."

Powerdrill Bloom's pages do not name Gemini 4 Argon as a model it runs. The two can work side by side. Use Argon for long reading and drafting, and use a file-first workspace where each figure must trace back to a source.

Does your proposal start from a spreadsheet or a PDF? You can try Powerdrill Bloom on one of your files and turn the findings into slides.

Frequently asked questions

What is Gemini 4 Argon?

Gemini 4 Argon is Google's frontier reasoning model, generally available since October 13, 2026. Google built it for complex coding, agentic, math, and science tasks. It also targets knowledge work such as finance and legal. Developers call it as gemini-4-argon in the Gemini API.

When was Gemini 4 Argon released?

Google announced it on September 30, 2026, for a set of trusted cyber defenders. The general release followed on October 13, 2026. That day it reached the Gemini API, AI Studio, Google Antigravity, the Gemini app, and AI Mode in Search.

Who can use Gemini 4 Argon, and is it free?

In the Gemini app, Gemini Apps Help lists Argon for Google AI Pro and Ultra plans. In the API, the pricing page's free tier column reads "Not available" for Argon. Paid API rates for each usage mode are on Google's pricing page.

Can Gemini 4 Argon make presentations?

Yes, through tasks in the Gemini app. Google's demo has Argon turn a folder of project files into a proposal deck for a client. Gemini Apps Help lists .pptx among the file types a task can create. Check every number on the slides against your source files.

How does Gemini 4 Argon compare with Claude Opus 5.5 and GPT-6?

Anthropic lists Claude Opus 5.5 at $4 input and $20 output per million tokens. OpenAI lists GPT-6 Astra at $10.00 and $50.00, and GPT-6.1 Sol at $2.00 and $10.00. Google lists Argon's API rates on its own pricing page. Test each model on your own documents before you choose.

Sources: Google, Gemini 4 Argon: Built to solve your hardest problems · Google, Introducing Gemini 4 Argon · Google, Gemini 4 Argon comes to the Gemini app · Google, Gemini 4 Argon is now available for developers · Google AI for Developers, Gemini API release notes · Google AI for Developers, Gemini 4 Argon · Google AI for Developers, What's new in Gemini 4 Argon · Google AI for Developers, Gemini API pricing · Google DeepMind, Gemini 4 Argon model card · Gemini Apps Help, Manage your tasks and workflows · Gemini Apps Help, Create and manage schedules for tasks · Gemini Apps Help, About Gemini models · Google One, Google AI plans · Anthropic, Introducing Claude Opus 5.5 · Claude Platform Docs, Models overview · OpenAI, API pricing. Details read from vendor pages on October 14, 2026.