Last updated: October 2026
Key Takeaways
- Gemini 4 Argon is public but not free: it requires Google AI Pro or Ultra in the Gemini app and AI Mode, or the paid Gemini API, and Google has named no free tier or date for one.
- AI Pro gets Argon in the app only. It is not included in AI Studio, the API, or Antigravity, and the API's launch price of $2 and $10 per million tokens doubles on December 1.
- Only the paid API keeps your prompts out of Google's training by default, and Argon has no downloadable weights. Google's open-weight Gemma 4 is the version you can own.
No. Gemini 4 Argon went public on October 13, 2026, but Google offers no free tier and has named no date for one. In the Gemini app and AI Mode in Search, it requires Google AI Pro or Ultra. Developers must use the paid Gemini API, at $2 per million input tokens and $10 per million output tokens through November 30.
Every figure below was checked against Google's own pages on October 13, 2026. The rollout is still moving, and this page updates with it.
Who Can Use Gemini 4 Argon Now?
Anyone with Google AI Pro, Google AI Ultra, or a paid Gemini API account can use Gemini 4 Argon as of October 13, 2026, per Google's announcement. Free accounts and Google AI Plus are left out.
| Account | Gemini app and AI Mode | AI Studio and Antigravity | Notes |
|---|---|---|---|
| No plan (free) | No | No | Auto model selection since October 9, mostly Flash-Lite |
| Google AI Plus | No | No | Plans page lists Gemini 3.6 Flash |
| Google AI Pro | Yes, rolling out | No | Google's API pricing page excludes the Pro plan |
| Google AI Ultra (both tiers) | Yes | Yes | 5x or 20x Pro's usage limits, depending on tier |
| Paid Gemini API | Not included | Yes, in AI Studio with billing on | Billed per token; model ID gemini-4-argon |
| Business and enterprise | Yes, in Gemini for business on eligible plans | Through the API and Antigravity | Also offered on Google Cloud |
| Selected scientists | No | Antigravity's top tier, free for 6 months | The only free route; up to 10,000 places, by application |
Checked October 13, 2026 against Google's announcement, developer post, Gemini plans page, and Gemini API pricing page.
Does Google AI Pro include Gemini 4 Argon?
Yes in the Gemini app and AI Mode in Search, and no everywhere else. Google's plans page lists Argon under AI Pro with a "rolling out" label, so it may not reach every Pro account on day one. Google's API pricing page excludes the Pro plan, and its developer post limits subscriber access in AI Studio and Antigravity to AI Ultra. We confirmed Argon on an AI Ultra account on October 13 but have not yet confirmed it on Pro.
What free Gemini users get instead
Free users get Google's smaller models, not Argon. Since October 9, accounts without a Google AI plan use Auto, which sends most prompts to Flash-Lite and routes harder ones to Flash or Pro, per Google's help page on model access.
How Much Does Gemini 4 Argon Cost?
Gemini 4 Argon costs $19.99 a month through Google AI Pro, $99.99 or $199.99 a month through the two Google AI Ultra tiers, or a per-token rate on the paid Gemini API, per Google's plans and pricing pages on October 13, 2026. Usage limits are compute-based, as our report on the Ultra tier split explains. On the API, the launch price runs through November 30 and doubles on December 1.
| Per 1M tokens | Through November 30, 2026 | From December 1, 2026 |
|---|---|---|
| Input | $2.00 | $4.00 |
| Output, including thinking tokens | $10.00 | $20.00 |
| Cached input | $0.10 | $0.20 |
| Batch input | $1.00 | $2.00 |
| Batch output | $5.00 | $10.00 |
From Google's Gemini API pricing page, October 13, 2026. Caching is implicit only; Google lists no explicit caching for Argon. Grounding with Google Search includes 5,000 free requests a month, then costs $14 per 1,000.
A large batch job sent before December 1 gets the lowest price Argon has announced. Google's Gemini API terms reserve the API for professional or business use.
How Argon's price compares
Argon's launch price is low but not unique: Claude Sonnet 5.5 and GPT-6.1 Sol list at the same $2 and $10, per Anthropic's pricing page and OpenAI's pricing page. At its December 1 rate, Argon matches Claude Opus 5.5.
| Model | Input per 1M tokens | Output per 1M tokens |
|---|---|---|
| Gemini 4 Argon, through November 30 | $2 | $10 |
| Gemini 4 Argon, from December 1 | $4 | $20 |
| Claude Sonnet 5.5 | $2 | $10 |
| GPT-6.1 Sol | $2 | $10 |
| Claude Opus 5.5 | $4 | $20 |
| GPT-6 Astra | $10 | $50 |
Short-context list prices on October 13, 2026, before batch and cache discounts. Claude Fable 5.1 lists at the same price as GPT-6 Astra.
The 1-million-token figure, corrected
Google now describes one 1-million-token budget shared by input and output, reached through a continuation protocol in the API. VentureBeat reports that long output needs API configuration and is not the Gemini app's default.
Correction, October 13, 2026: an earlier version of this page said Argon could write up to 1 million tokens in a single response. Google's current documentation describes a combined input and output limit.
Paying for Pro or Ultra Doesn't Buy API-Grade Privacy
Gemini app chats on a paid AI Pro or Ultra plan follow the same terms as free chats: Google can train on them unless you turn off Keep Activity. Google's Gemini Apps Privacy Hub, updated October 13, makes no distinction between free and paid plans. Only the paid API keeps prompts out of product improvement automatically.
| Where you use it | Used to improve Google's AI? | How long it is kept |
|---|---|---|
| Gemini app, Keep Activity on | Yes, including model training; human reviewers read some chats | 18 months by default; reviewed chats up to 3 years |
| Gemini app, Keep Activity off or temporary chat | No, unless you send feedback; reviewers can still see chats for safety | 72 hours; any chat a reviewer sees, up to 3 years |
| Gemini API, unpaid tier | Yes; human reviewers may read prompts and responses | Not specified in the API terms |
| Gemini API, paid tier | No | Logged for a limited period for abuse checks and legal disclosures |
| Gemma 4 on your own hardware | No; nothing leaves your machine | As long as you choose to keep it |
From Google's Gemini Apps Privacy Hub (updated October 13, 2026) and Gemini API Additional Terms (updated April 28, 2026). Work and school accounts may be subject to different terms.
How to keep Argon chats out of training
On Pro or Ultra, turn off Keep Activity in Gemini Apps Activity or use a temporary chat, and skip the feedback buttons. Google still keeps those chats for 72 hours and keeps any chat a reviewer sees for up to three years. Our guide to turning off AI features covers Google's AI Overviews and Gemini on Android.
Own the hardware your internet runs through
Keep Activity has an off switch. The remote access on a gateway your ISP rents you usually does not.
Rented gateways typically run TR-069 remote-management software, which lets the provider administer the router at the edge of your home network. Owning your router moves that control to you, and owning your modem ends the monthly rental fee. Our guide to owning your modem covers what changes.
The Argon You Pay For Ships With Cyber Guardrails
The public version of Argon blocks offensive security work. Google Cloud's launch post says it refuses tasks such as penetration testing and exploitation, while defensive work goes through Google's CodeMender agent or your own tooling. Only vetted defenders in the Fairwind Program, now expanding to hundreds more partners, get Argon without them. Google's safety post says monitoring can stop a response partway through.
OpenAI and Anthropic draw similar lines, as our Opus 5.5 pricing breakdown explains. The capability you pay for is set by the vendor's view of who you are, and the vendor can change it, sometimes silently, as the backlash over Claude Fable 5's silent safeguards showed. Only a model running on your own hardware is beyond a vendor's reach.
Is Gemini 4 Argon Worth Paying For?
If you already pay for AI Pro, try it; it costs nothing extra. Independent scores put Argon at or near the top, but not alone there. Vals AI and Arena rank it first, while Artificial Analysis scores it level with GPT-6 Astra and Claude Fable 5.1 and behind two Anthropic models.
| Evaluator | Measure | Gemini 4 Argon |
|---|---|---|
| Vals AI | Vals Index (finance, legal, tax, and coding tasks) | 68.9%, ranked first, ahead of Claude Sonnet 5.5 and Opus 5.5 |
| Arena | Text Arena (human preference votes) | 1,525 points, ranked first; rating marked preliminary |
| Artificial Analysis | Intelligence Index | 53, tied with GPT-6 Astra and Claude Fable 5.1; Claude Opus 5.5 scores 58 and Sonnet 5.5 scores 56 |
Arena and Artificial Analysis checked October 13, 2026. Vals standing as of October 12 and the Opus 5.5 and Sonnet 5.5 index scores as reported by VentureBeat.
Google's own chart is split on coding: Argon leads DeepSWE v1.1 at 77.9% to Claude Opus 5.5's 74.2% but trails Opus 5.5 on Terminal-Bench 4.0, 57.4% to 66.4%. Its widest leads are in legal agent work, long-context retrieval, and business automation.
The read: Pro subscribers should test Argon on long documents and legal or finance agent work. Don't upgrade to Ultra for Argon alone unless you need AI Studio or Antigravity, and run large API batch jobs before December 1.
Can You Run Gemini 4 Argon Locally?
No. Google has not released Argon's weights and has announced no open or on-device version, so there is no download and no self-hosted build. Sites offering a free Argon download are not Google; give them no login or API key.
Google's open-weight line is Gemma 4, released under Apache 2.0 in five sizes, per the Gemma 4 model card: E2B and E4B for phones and small machines, 12B, a 26B mixture-of-experts model with about 4 billion active parameters, and 31B. Our Gemma 4 launch coverage has the background, and our Gemini 3.6 Flash reality check maps each size to the hardware that runs it.
Gemma 4 is not a frontier model, but for drafting, summaries, document questions, and everyday coding it runs on your terms, and nothing you type leaves the machine. Our models-by-VRAM guide ranks open models at every memory tier, and our local AI hardware guide covers the machines.
The same logic applies to your modem. A rented modem is a monthly fee for hardware you never own, and buying one outright ends that fee. For cable internet, DOCSIS 3.1 is the newest standard sold at retail; check your ISP's approved list before you buy.
Frequently Asked Questions
Is Gemini 4 Argon free?
No. Since October 13, 2026, Gemini 4 Argon has been available to Google AI Pro and Ultra subscribers in the Gemini app and AI Mode in Search, and to developers on the paid Gemini API. Free accounts, Google AI Plus, and the API's free tier do not include it, and Google has named no date for free access. The only free route is a program for up to 10,000 scientists.
Does Google AI Pro include Gemini 4 Argon?
Yes, in the Gemini app and AI Mode in Search, where Google lists it as rolling out to Pro subscribers. Pro does not include Argon in Google AI Studio, the Gemini API, or Google Antigravity: Google's API pricing page excludes the Pro plan, and subscriber access in AI Studio and Antigravity is limited to Google AI Ultra.
Can I use Gemini 4 Argon in Google AI Studio for free?
No. Google's Gemini API pricing page lists no free tier for Argon. In AI Studio you can use it with a billed Gemini API account or as a Google AI Ultra subscriber. Free AI Studio users and Google AI Pro subscribers do not get Argon there.
How much does the Gemini 4 Argon API cost?
Through November 30, 2026, Gemini 4 Argon costs $2 per million input tokens and $10 per million output tokens on the paid Gemini API. From December 1, it costs $4 and $20. Thinking tokens bill as output, cached input costs 95% less, and batch requests cost half, per Google's pricing page on October 13.
How do I use Gemini 4 Argon?
Subscribe to Google AI Pro or Ultra, then open the Gemini app or AI Mode in Google Search with that account; on Pro it is still rolling out. Developers call it on a billed Gemini API key with the model ID gemini-4-argon, and AI Ultra subscribers can also use it in Google AI Studio and Antigravity.
Will free Gemini users get Gemini 4 Argon?
Google has not said. Its announcement names no free tier and no date. Since October 9, 2026, free Gemini accounts use Auto, which sends most prompts to Flash-Lite. For a capable model at no ongoing cost today, Google's open-weight Gemma 4 runs free on hardware you own.
Is Gemini 4 Argon the same as Gemini 4 Pro?
Google has not announced a Gemini 4 Pro. Argon is the only Gemini 4 model in Google's API documentation, and Google says more are coming. Business Insider reported on October 10 that staff were testing internal versions codenamed Barium and Carbon, per VentureBeat; Google declined to comment, so treat "Gemini 4 Pro" claims as unconfirmed.
Can I run Gemini 4 Argon on my own computer?
No. Google has not released Argon's weights, and any site offering a free download is not Google. Google's open-weight family is Gemma 4, licensed Apache 2.0 in sizes from E2B to 31B, which runs on phones, laptops, mini PCs, and consumer GPUs. It is less capable than Argon, but it runs entirely on hardware you own.
