Google's Gemini 4 Carbon is out, but mostly for paying subscribers and developers
Google's strongest AI model opened up on November 10. In the Gemini app it's only on the top AI Ultra plan, and in Search it now also reaches AI Pro subscribers.
Developing story · Gemini 4 Carbon's rollout- Google made Gemini 4 Carbon, its most powerful AI model, generally available on November 10, 20261,6.
- In the Gemini app it's only for Google AI Ultra subscribers, a plan that starts at $100 a month in the US1,5. AI Mode in Search added AI Pro subscribers by November 1625.
- As of November 18, independent tester Artificial Analysis ranks it third, behind Anthropic's Claude Fable 5.2 and OpenAI's GPT-6.2 Astra8,23.
- Developers pay $2/$10 per million tokens until November 30, then $4/$201.
Should you care?
2/5 Matters if you pay for Google AI or code
How hot right now
WarmWide release on November 10 drew coverage; nothing big new this week
Fire meter checked Nov 19, 2026
Hype check
Benchmark charts are loud, but most Gemini users stay on Gemini 4 Argon or lower, and Google still points most coding and knowledge work to Argon7.
We'll check back on Feb 14, 2027: did it live up to the hype?
What it costs where you live
Who is this for?
- ✕Free Gemini app usersSkip itCarbon is not offered on the free tier; in the app it is limited to AI Ultra subscribers1.
- ~Google AI Pro subscribersMaybeYou can now pick Carbon in AI Mode in Search, Search Engine Roundtable reports, but the Gemini app still gives you Argon7,25.
- ✓Google AI Ultra subscribersFor youCarbon is rolling out to you in the Gemini app, AI Mode in Search and Antigravity1.
- ✓Developers building with AI APIsFor youIt is in the paid Gemini API at an introductory $2/$10 per million tokens through November 301.
- ~Security teamsMaybeA cyber version with fewer guardrails is available, but only to approved members of Google's Fairwind Program1,5.
Google's smartest AI is here, but mostly behind a paywall
Google released Gemini 4 Carbon, its most powerful AI so far. Unless you pay for Google's AI plans or write software, your Gemini app and Search won't change.
Google makes an AI assistant called Gemini. It lives in the Gemini app on Android phones and iPhones, and inside Google Search. Every few months Google swaps in a smarter "brain" behind it. The newest one is called Gemini 4 Carbon, and it became widely available on November 10, 20261,6.
So will my Gemini get smarter?
Probably not. In the Gemini app, Carbon is only switched on for people who pay for Google AI Ultra, Google's top plan, which starts at $100 a month in the US5. Ultra subscribers can pick it from a menu in the Gemini app and in AI ModeJargon busterA chat-style section of Google Search where an AI writes the answer., the chat-style part of Google Search1,7.
If you pay for the cheaper Google AI Pro plan, you may now see it in AI Mode too. Search Engine Roundtable reported on November 16 that Google had opened it to Pro subscribers there, which a Google search executive confirmed25. In the Gemini app, Pro subscribers keep the modelJargon busterThe AI "brain" behind a chatbot; companies swap in new models to make the assistant smarter. they already had, Gemini 4 Argon7. If you use Gemini for free, nothing changes.
Think of it like a restaurant adding a chef's tasting menu. It's impressive, but most diners are still ordering from the regular menu, and that menu didn't get worse.
What is it good at?
Google says Carbon is built for long, complicated jobs: writing and fixing big chunks of software, scientific research, and finding security holes in computer code2,5. Google itself still suggests the older Argon model for most everyday coding and knowledge work7.
Is it better than ChatGPT or Claude?
It's close at the top. An independent testing company, Artificial Analysis, first ranked Carbon number one on its scoreboard, then updated its tests8,10. As of November 18 it puts Carbon third: behind Anthropic's Claude Fable 5.2 and OpenAI's GPT-6.2 Astra, which came out on November 17, and one point ahead of Anthropic's Claude Opus 5.58,23,24. The scores are close enough that for chatting, writing emails or planning a trip, you likely wouldn't feel a difference.
What should I do?
Nothing, unless you already pay. Ultra subscribers can look for "4 Carbon" in the Gemini app's model menu7, and Pro subscribers can check the model menu in AI Mode25. Save it for your hardest questions.
What should you do?
A gated debut, a tiered rollout and an aggressive API price
Carbon started with security defenders, reached developers and Ultra subscribers two weeks later and Pro users in Search after that. It launched at a lower APIJargon busterA way for programmers to plug an AI model into their own apps and pay per use. price than Claude Fable 5.2 and Opus 5.5, and its rank depends on who's grading.
Carbon didn't start with consumers. Google announced it on October 28 as its "next frontier modelJargon busterThe AI "brain" behind a chatbot; companies swap in new models to make the assistant smarter.," at first giving it only to cybersecurity defenders in its gated Fairwind ProgramJargon busterGoogle's gated program that gives approved cybersecurity defenders early or less-restricted model access.2,14. It's the second Gemini 4 model, after Argon reached the Gemini app in October12.
The tiering, in detail
- Ultra subscribers also get it in AntigravityJargon busterGoogle's AI coding tool, where AI agents help write and fix software., Google's coding tool1. In AI ModeJargon busterA chat-style section of Google Search where an AI writes the answer. it's limited to subscribers 18 and older in supported countries, with daily usage limits6.
- Pro subscribers got Carbon in AI Mode by November 16, which Google's Rajan Patel confirmed to Search Engine Roundtable25. Google hasn't given a timeline for Pro in the Gemini app7, where the model picker now labels Argon "Code & analysis" and Carbon "Most complex work," and Gemini 3.1 Pro is gone7.
- Developers and businesses get it through the paid Gemini APIJargon busterA way for programmers to plug an AI model into their own apps and pay per use. in Google AI Studio and Gemini for business1.
- Security teams get a "Gemini 4 Carbon for cyber" variant with fewer guardrails, which stays behind a Fairwind allowlist1,5.
Not everyone was happy with the tiering. A thread on Google's own developer forum asked for Carbon in AI Pro across Antigravity, the Gemini app and Search15; Search is the first of those to happen. One catch there: AI Mode's header no longer names the model you picked, which Google says is intended25.
Whose scoreboard?
Google's launch table shows Carbon with the top score in 15 of 17 rows9. Those are Google's own numbers, and VentureBeat noted that OpenAI kept the lead on one software-engineering test and Anthropic on one computer-use test9. The independent ranking above tells a more modest story, and on Artificial Analysis's coding-agent index Carbon trails Claude Code running Opus 5.510.
The trade-off worth knowing: Carbon makes things up more often than Argon. Its hallucinationJargon busterWhen an AI states something false with confidence. rate on the AA-Omniscience test is 11%, versus 4% for Argon11. Our read: that makes Google's advice to keep Argon for everyday work look sensible.
The price play
Carbon launches at Argon's introductory API price, $2 per million input tokens and $10 per million output tokens, until November 30. From then on it's $4 and $201. VentureBeat calculates that's about a quarter of Fable 5.2's standard rates during the introductory period, and half after the increase8. Artificial Analysis puts the cost of running its full test suite at $7.76 per task for Carbon versus $10.04 for Opus 5.510.
What to watch
Whether Carbon reaches Pro subscribers in the Gemini app and Antigravity, and whether developers stick around once the API price doubles on December 1.
By the numbers
How we got here
- Oct 13, 2026Gemini 4 Argon reaches the Gemini app for Pro and Ultra subscribers12
- Oct 28, 2026Google announces Gemini 4 Carbon, limited to Fairwind cyber defenders2,14
- Nov 10, 2026Carbon goes generally available to paying developers and Ultra subscribers1,6
- Nov 16, 2026Carbon reported live in AI Mode for AI Pro subscribers25
- Nov 17, 2026OpenAI releases GPT-6.2 Astra, which edges Carbon on Artificial Analysis's index23,24
- Nov 30, 2026 · coming upIntroductory API price ends1
- Dec 1, 2026 · coming upAPI price rises to $4 / $20 per 1M tokens1,6
Model card details: output budgets, caching, benchmarks and caveats
How Carbon's shared context budget and implicit caching work, Google's self-reported agenticJargon busterDescribes AI that carries out multi-step tasks on its own, like running code and checking results. and security benchmarks, and the full Artificial Analysis numbers behind the rankings above.
Google's modelJargon busterThe AI "brain" behind a chatbot; companies swap in new models to make the assistant smarter. card positions Carbon for long-horizon agenticJargon busterDescribes AI that carries out multi-step tasks on its own, like running code and checking results. work across knowledge work, engineering, cybersecurity and scientific reasoning3. Besides the APIJargon busterA way for programmers to plug an AI model into their own apps and pay per use., it's on Google Cloud's Gemini Enterprise Agent Platform, together with the cyber-permissive variant2,13.
Context and output budget
| Spec | Value |
|---|---|
| Default output per request | 65,536 streaming / 32,768 non-streaming4 |
| Max output (long decode) | Up to the full window minus input length4 |
| Caching | Implicit only, for shared prefixes of 4,096+ tokens; no explicit caching4 |
Because input and output share one window, the output budget counts thinking tokens and function calls, not just the visible answer4. Google Cloud's pricing page lists extra tiers beyond the two API rates above that we couldn't fully reconcile, so check it for your deployment.
Google's self-reported benchmarks
These come from Google's launch post and aren't independently reproduced2,9:
| BenchmarkJargon busterA standardized test used to compare AI models; companies often run and report their own. | Argon | Carbon |
|---|---|---|
| Terminal-Bench 4.0 (AA) | 57.1% | 65.2%9 |
| Agents' Last Exam v1.1 | 69.0% | 77.4%9 |
| SWE-sweep | 4.3% | 9.6%9 |
| SEC-bench Pro (find vulns) | n/a | 88.2%9 |
| ExploitGym (exploit) | n/a | 54.8%9 |
| CWE-bench (patch) | n/a | 71.0%9 |
On FrontierSWE 2, Carbon's 61.2% trails GPT-6 Astra (65.5%) and Claude Opus (62.3%) in the same table9.
Independent results (Artificial Analysis)
| Test | Carbon | Comparison |
|---|---|---|
| Intelligence Index v5.0, at launch | 5910 | Opus 5.5 58, GPT-6 Astra 5710 |
| Same index after the regrade, high reasoning | 608 | Fable 5.2 max 658, GPT-6.2 Astra max 6223, Opus 5.5 max 59, GPT-6 Astra max 578 |
| Coding Agent Index | 5910 | Claude Code with Opus 5.5 max: 6210 |
| AA-Omniscience v1.1 | 5411 | Opus 5.5 48, Argon 4611 |
| Cost to run the index, per task | $7.76 at high10 | Opus 5.5 max $10.0410, Fable 5.2 max $26.548 |
The hallucinationJargon busterWhen an AI states something false with confidence. rate quoted above is incorrect answers divided by incorrect, partial and abstained answers, not by all answers11.
Caveats and open questions
Scores move when evaluators update their suites, as Artificial Analysis's own regrade shows8, and new releases reshuffle the order: GPT-6.2 Astra passed Carbon on November 17, a week after Carbon's general release23,24. The price per tokenJargon busterA small chunk of text, roughly three-quarters of a word, used to measure and bill AI usage. is low, but Carbon spends many tokens on reasoning, so cost per task matters more than list price. The cyber variant's guardrail differences aren't publicly documented beyond it being gated behind Fairwind1.
The facts
Hover or tap anything underlined to see what it means.
- Made by
- Google DeepMind3
- Model name
- Gemini 4 Carbon (built on Gemini 4 Argon)3
- Announced
- October 28, 2026, cyber defenders only2
- Consumer access
- Gemini app and AntigravityJargon busterGoogle's AI coding tool, where AI agents help write and fix software.: AI Ultra only1. AI ModeJargon busterA chat-style section of Google Search where an AI writes the answer.: Ultra, plus AI Pro since mid-November1,25
- Ultra price
- From $100 a month in the US5,16; local prices from Google's country pages18,19,20,26,27 and HelenTech for Japan22
- Free tier
- Not included1
- Intro API price
- $2 in / $10 out per 1M tokens, to Nov 301
- API price later
- $4 in / $20 out per 1M tokens1
- Context window
- 1,048,576 tokens, shared in/out4
- Inputs / output
- Text, image, audio, video in; text out4
- Knowledge cutoff
- March 20263
Jargon buster
- Model
- The AI "brain" behind a chatbot; companies swap in new models to make the assistant smarter.
- Google AI Ultra / Pro
- Google's paid AI subscription plans; Ultra is the most expensive tier.
- AI Mode
- A chat-style section of Google Search where an AI writes the answer.
- Antigravity
- Google's AI coding tool, where AI agents help write and fix software.
- API
- A way for programmers to plug an AI model into their own apps and pay per use.
- Token
- A small chunk of text, roughly three-quarters of a word, used to measure and bill AI usage.
- Context window
- How much text, images or audio the model can consider at once, measured in tokens.
- Benchmark
- A standardized test used to compare AI models; companies often run and report their own.
- Hallucination
- When an AI states something false with confidence.
- Fairwind Program
- Google's gated program that gives approved cybersecurity defenders early or less-restricted model access.
- Agentic
- Describes AI that carries out multi-step tasks on its own, like running code and checking results.
- Knowledge cutoff
- The date after which the model has no built-in knowledge of the world.
Gemini 4 Carbon's rollout
Google's strongest AI model is opening up step by step, while rivals keep leapfrogging it.
- Oct 28, 2026Google announces Gemini 4 Carbon, at first only for cyber defenders Google
- Nov 10, 2026Carbon opens to paying developers and AI Ultra subscribers You're reading this
- Nov 16, 2026Carbon reported live in Search's AI Mode for AI Pro subscribers Search Engine Roundtable
- Nov 17, 2026OpenAI's GPT-6.2 Astra edges past Carbon on an independent ranking Artificial Analysis
- Dec 1, 2026 · coming upDeveloper prices rise to $4 / $20 per million tokens Google
- Whether Carbon reaches the Gemini app for AI Pro subscribers
- The developer price rise on December 1
Sources25 pages · Google, Google DeepMind, Google Cloud and moreShowHide
The small numbers in the text point to these pages. Everything is written in our own words. How we work · Spotted a mistake? Tell us
- 1Gemini 4 Carbon is now available · Google
- 2Gemini 4 Carbon: our next frontier model · Google
- 3Gemini 4 Carbon model card · Google DeepMind
- 4Gemini 4 Carbon (Gemini Enterprise Agent Platform docs) · Google Cloud
- 5Google's Gemini 4 Carbon is now widely available around the world · Engadget
- 6Google rolls out Gemini 4 Carbon to AI Ultra subscribers · Android Authority
- 7Gemini 4 Carbon rolling out to Gemini app for AI Ultra, Argon remains · 9to5Google
- 8Google Gemini 4 Carbon is now publicly available, and it undercuts Claude Fable 5.2 and Opus 5.5 on API price · VentureBeat
- 9Google unveils Gemini 4 Carbon, retaking the world's no. 1 spot on updated, harder benchmarks, but access is limited · VentureBeat
- 10Gemini 4 Carbon is the new #1 on the Artificial Analysis Intelligence Index v5 · Artificial Analysis
- 11AA-Omniscience v1.1: Knowledge and Hallucination Benchmark · Artificial Analysis
- 12Google's Gemini 4 Argon is now available in the Gemini app and AI Mode in Search · Engadget
- 13Gemini Enterprise Agent Platform release notes · Google Cloud
- 14Google announces Gemini 4 Carbon with 'significant gains across coding and agentic tasks' · 9to5Google
- 15Unfair Tier Gating: Include Gemini Carbon Access in Google AI Pro · Google AI Developers Forum
- 16Google AI Pro & Ultra plans · Google
- 18Google AI plans (France) · Google
- 19Google AI plans (Australia) · Google
- 20Google AI plans (Canada) · Google
- 22Google AI Ultra が月額 14,500 円から利用可能に。上位プランも 32,000 円に値下げ · HelenTech
- 23GPT-6.2 Astra Models - Intelligence, Performance & Price Comparison · Artificial Analysis
- 24OpenAI's GPT-6.2 Astra trails Claude Fable 5.2 by three points on an independent AI index at a fraction of the cost per task · R&D World
- 25Google AI Mode Drops Gemini Label & Rolls Out Gemini 4 Carbon To Pro Subscribers · Search Engine Roundtable
- 26Google AI plans (UK) · Google
- 27Google AI plans (India) · Google
Read next
AI · ★ Saved
Gemini will now hold on to your "off the record" chats for about two months, not 72 hours
Google will keep Gemini chats for about two months even when you've turned history off or used a Temporary Chat. In return, you get a separate switch to stop Google training its AI on what you type.
AI · ★ Saved
OpenAI's GPT-6.2 Astra and Sol are here, but not for free users or regular ChatGPT chats
OpenAI launched GPT-6.2 Astra and GPT-6.2 Sol, two upgraded AI models for paying ChatGPT users. They only show up in ChatGPT's Work and Codex modes, so free users and ordinary chats don't change.
AI · ★ Saved
Microsoft's draft AI rulebook says people stay in charge, but only for its own models
Microsoft's AI team published a draft set of rules saying its own AI models must always let people stop, correct or redirect them. It's a promise for future models, so nothing changes in Copilot or Office today.