Between 21 July and 28 October 2026, Google released or announced 16 Gemini-branded models, by our count of its Gemini API release notes and its own blog posts. The latest, Gemini 4 Carbon, was announced on 28 October 2026, the day Alphabet reported its third-quarter results.
On Google’s Gemini API pricing page, Gemini 4 Argon costs $2.00 per million input tokens and $10.00 per million output tokens through 30 November 2026, then $4.00 and $20.00 from 1 December. Google says Carbon “will launch at the same introductory price” as Argon, “$2 per million input tokens and $10 per million output tokens.”
Business Pill · Usage-Based Pricing
An evergreen Business Engineer explainer on usage-based pricing. It explains the general idea only and does not discuss Google’s models or prices.
The key insight: By our arithmetic, Google launched 16 Gemini-branded models in 99 days, and Carbon’s stated introductory price of $2 and $10 per million tokens is half of Claude Opus 5.5’s list price.
Google announced Gemini 4 Argon on 30 September 2026, saying it was “rolling out to a set of trusted cyber defenders through our Fairwind Program.” On 13 October, Google said Argon “is now globally available today for developers and enterprises via the API in AI Studio and Gemini platform.” The release notes list it as generally available that day.
Google’s Carbon post calls it “our strongest model with state-of-the-art cybersecurity and long horizon capabilities.” Google says Carbon “ranks #1 on the recently updated v5.0 of Artificial Analysis Intelligence Index at 59.”
On access, Google says: “Starting this week, we will make Carbon available to a subset of cyber defenders and will gradually phase access to additional trusted partners through our Fairwind Program.” On the earnings call, Sundar Pichai said Google is “working on distilled versions of our Carbon models.”
Gemini-branded models Google released or announced between 21 July and 28 October 2026, by date, grouped as Flash family, voice, media and robotics, and Gemini 4 frontier models. Variants are counted separately; some models had earlier previews. Sources: Gemini API release notes (ai.google.dev); Google and Google DeepMind blog posts. The count of 16 and the 99-day span are our arithmetic.
What Was Posted on X
Google is rolling out Gemini 4 Carbon to selected cybersecurity defenders through Fairwind at $2/1M input and $10/1M output tokens, rising to $4 and $20 later (@carlfranzen / VentureBeat)
(Visit Techmeme dot com for the link and full context!)
Techmeme, citing VentureBeat’s Carl Franzen, says Google is rolling out Gemini 4 Carbon to selected cybersecurity defenders through Fairwind at $2 per million input and $10 per million output tokens, rising to $4 and $20 later.
The Flash Family
The release notes list Gemini 3.6 Flash and Gemini 3.5 Flash-Lite as generally available on 21 July 2026, Gemini 3.7 Flash on 13 August and Gemini 3.8 Flash on 2 September. Google describes 3.8 Flash as “our most intelligent Flash model.”
The same day, Google introduced Gemini 3.8 Flash Cyber, which it calls “our most capable cybersecurity model with frontier-level performance in vulnerability detection and automated patching, available to trusted defenders through our new Fairwind Program.”
On 8 October, the release notes say gemini-3.7-flash was deprecated and requests are “automatically routed to gemini-3.8-flash.”
Voice, Media and Robotics Models
The other nine cover robotics, speech, video and images. The release notes list them by date:
30 July: Gemini Robotics ER 2, in public preview.
26 August: Gemini 3.5 Transcribe and Gemini 3.5 Transcribe Live.
27 August: Gemini Omni 1.1 Flash, the generally available version of Google’s video generation and editing model.
15 September: Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking.
6 October: Gemini Nano Banana 2.1, an image generation and editing model.
Google DeepMind also announced Genie 4, a world model that is not Gemini-branded, on 27 October; we covered it in Genie 4 is Google’s first real-time world model with sound. Its post says Google has “started rolling Genie 4 out as a research preview for trusted testers.” The release notes also list Lyria 3.5, a music model, on 3 September. Neither is in our count of 16.
Where the Models Show Up
In his third-quarter remarks, Pichai said: “The Gemini app is now at one billion users and is the fastest-growing product in Google’s history.” He said agentic features would extend “beyond Search and the Gemini app to Chrome, Android, and our Workspace products as well.”
On the call, Alphabet’s CFO, Anat Ashkenazi, said “Google One is our fastest-growing subscriptionproduct, driven by demand for AI plans.” Google’s 13 October post said Argon was available to Google AI Ultra and Pro subscribers in the Gemini app and AI Mode in Search.
For companies, Thomas Kurian, CEO of Google Cloud, said on the call that “Gemini Enterprise now includes the Gemini agent that is deeply integrated across our entire cloud portfolio.” He also named “our agent compute engine which is specifically designed for AI agents,” alongside TPU systems and Axion CPUs.
How the List Prices Compare
Per million tokens, standard paid tier, on the pricing pages we read on 28 October 2026:
Google Gemini 3.8 Flash and Gemini 3.6 Flash: $0.75 input and $3.75 output through 31 December 2026, then $1.50 and $7.50 from 1 January 2027.
Google Gemini 3.5 Flash-Lite: $0.30 input and $2.50 output.
Anthropic Claude Opus 5.5: $4 input and $20 output. Claude Sonnet 5.5: $2 and $10. Claude Fable 5.1: $10 and $50.
OpenAI GPT-6 Astra: $10.00 input and $50.00 output. GPT-6.1 Sol: $2.00 and $10.00. GPT-6.1 Terra: $1.00 and $5.00.
By our arithmetic, Carbon’s stated introductory price is half the list price of Claude Opus 5.5 and one-fifth that of GPT-6 Astra and Claude Fable 5.1. From 1 December, Argon’s listed price of $4.00 and $20.00 matches Opus 5.5.
The Release Pace
By our arithmetic, 16 launches across the 99 days from 21 July to 28 October is one every six days or so. The launches fell on 11 separate dates, and every month from August to October brought at least one new Flash or frontier model.
As we read the release notes, most of the 16 are not frontier models: nine cover robotics, speech, video and images, and five are Flash or Flash-Lite models.
As we read the pricing page, Argon’s $2.00 and $10.00 run through 30 November 2026, and the $0.75 and $3.75 for both 3.8 Flash and 3.6 Flash run through 31 December 2026.
By our arithmetic, the 16 launches fell on 11 separate dates between 21 July and 28 October.
Google, Gemini 4 Carbon announcement, 28 October 2026
“Starting this week, we will make Carbon available to a subset of cyber defenders”
Three Implications
FOR DEVELOPERS Google’s pricing page lists Argon at $4.00 and $20.00 per million tokens from 1 December 2026, up from $2.00 and $10.00.
FOR BUYERS COMPARING LABS On the pricing pages we read on 28 October, Claude Opus 5.5 lists at $4 and $20 and GPT-6 Astra at $10.00 and $50.00 per million tokens.
FOR CONSUMERS Pichai said the Gemini app is now at one billion users, and Google said Argon reached Google AI Ultra and Pro subscribers in the Gemini app and AI Mode.
The framework describes the token this way: “It is simultaneously the unit of cognition the model produces, the unit of compute the data center serves, the unit of price the lab charges, and the unit of value the enterprise extracts.”
Google’s catalogue prices each model per million tokens. The framework does not discuss these models or prices.
What Is Not Established
Our count includes only models whose names start with Gemini and that the release notes or Google’s blogs list between 21 July and 28 October. Variants such as Transcribe Live and Live Extended Thinking count separately, and Argon counts once, at its 30 September announcement. Some models, such as 3.6 Flash, had earlier preview versions.
Carbon does not yet appear on the Gemini API pricing page, and Google’s Carbon post does not say when its introductory price ends. Google does not publish usage by model, so the list does not show which models carry the most traffic.
This Week In The AI Supercycle
Execution Fault Lines, Concentrated Demand, and Geopolitical Firewalls
This week’s issue: the AI build is not short of demand for its parts. It is short of cheap money, and its end revenue rests on a narrow base of large, often venture-funded buyers.
Google launched 16 Gemini-branded models in 99 days, by our count from its own release notes and blog posts. Carbon’s stated introductory price of $2 and $10 per million tokens is half of Claude Opus 5.5’s list price, and Argon’s price is scheduled to double on 1 December.
Business Engineer Premium
Premium: Alphabet & Microsoft on the Map of AI
Cash conversion, the lengthening backlog, the arms-dealer turn at Layer 3: the full Business Engineer read of both prints.
95,000+ executives read Business Engineer for the AI strategy frameworks cited by ChatGPT, Claude, and Perplexity.
A note on sourcing. Models, dates and prices come from Google’s Gemini API release notes and pricing page and from Google and Google DeepMind blog posts; call quotations come from a live-caption transcript of Alphabet’s third-quarter 2026 earnings webcast of 28 October 2026; competitor prices come from Anthropic’s and OpenAI’s API pricing pages read on 28 October 2026. The count of 16, the day span and price ratios are our arithmetic; linked earlier pieces carry their own sources.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
