Skip to main content
Van Data Team

Current model · Google

Gemini 4 Carbon benchmarks, pricing and specs

Gemini 4 Carbon is a Google model released on October 28, 2026 that takes text and image input and produces text. It costs $2 per 1M input tokens and $10 per 1M output tokens. It scores 1528 Elo on Arena Text, #1 of 52 current models (Arena, measured October 28, 2026).

Per 1M input tokens
$2Listed today
Per 1M output tokens
$10Listed today
Arena Text
1528 Elo#1 of 52 current models

Data verified November 5, 2026 · Dataset 2026-11-05

Share this

Specs

Key facts

Provider
Google
Model family
Gemini 4
Release date
October 28, 2026
Status
Current
Access
limited access
API model id
Not reported
Context window
Not reported
Max output
Not reported
Input
text, image
Output
text
Task families
text, multimodal
Declared capabilities
reasoning, coding, agentic coding, image understanding, chart understanding

Charts

Against every current text model

Where Gemini 4 Carbon lands among the current models of its task, on every independent benchmark, on price and on context.

Where it stands

On Arena Text, Gemini 4 Carbon ranks #1 of 52 current text models. 29 of 55 list a lower price per 1M input tokens. Each bar is the share of current models it beats.

Gemini 4 Carbon among the current text models, one row per benchmark, price and context window.

Independent benchmarks

Arena Text

#1of 52

1,528 Elo · ahead of 100%

Arena WebDev

#4of 46

1,768 Elo · ahead of 93%

Price and context

Price per 1M input tokens

#30of 55

$2 · cheaper than 24%

Price per 1M output tokens

#35of 55

$10 · cheaper than 26%

Scores fromArena · 2026-11-03

Quality against price

Arena Text · measured October 28, 2026

Gemini 4 Carbon scores 1,528 on Arena Text at $4 (blended price per 1M tokens). The line joins the 5 current models no other beats on both price and score; hover any point to name it.

Quality against priceArena Text score against price for 44 text models, Gemini 4 Carbon marked.1,2501,3001,3501,4001,4501,5001,550$0.30$1$3$10MOST ATTRACTIVE QUADRANTMedian price $3Median 1,476Blended price per 1M tokens (3 input : 1 output), log scaleArena Text (Elo)Gemini 4 Carbon
Quality against price: the data
Arena Text score against price for 44 text models, Gemini 4 Carbon marked.
ModelProviderBlended price per 1M tokensArena Text (Elo)
Gemini 4 CarbonGoogle$41,528
Gemini 4 ArgonGoogle$41,520
Claude Opus 5.5Anthropic$81,509
Muse WonderMeta$81,507
Claude Fable 5Anthropic$201,504
Claude Fable 5.1Anthropic$201,504
Gemini 3.8 FlashGoogle$1.501,496
Muse Spark 1.3Meta$21,495
Claude Opus 5Anthropic$101,492
Kimi K3Moonshot Kimi$61,488
GPT-5.6 SolOpenAI$81,485
Gemini 3.6 FlashGoogle$31,483
GPT-5.5OpenAI$111,482
Claude Opus 4.8Anthropic$101,482
Qwen3.8-MaxAlibaba Qwen$31,482
GPT-6.1 SolOpenAI$41,481
GLM-5.3Z.ai$2.151,481
MiMo-V2.6-ProXiaomi MiMo$0.541,480
Claude Sonnet 5.5Anthropic$41,477
GLM-5.3-FlashZ.ai$0.241,477
GPT-6 AstraOpenAI$201,476
DeepSeek-V4.1-FlashDeepSeek$0.521,476
GPT-5.4OpenAI$5.631,475
Gemini 3 Flash PreviewGoogle$1.131,473
Grok 4.20xAI$1.561,472
GPT-5.6 TerraOpenAI$4.501,467
Grok 4.5xAI$31,465
DeepSeek-V4-ProDeepSeek$1.981,465
Claude Sonnet 5Anthropic$41,462
GPT-6 SolOpenAI$41,458
Gemini 3.5 Flash-LiteGoogle$0.851,456
Grok 4.6xAI$31,455
GPT-5.6 LunaOpenAI$0.451,453
MiMo-V2.6-FlashXiaomi MiMo$0.181,448
Step 5 PreviewStepFun$1.431,448
Grok 4.3xAI$1.561,442
GPT-6 LunaOpenAI$0.201,441
Grok 4.7xAI$31,441
Mistral Large 4Mistral AI$1.051,432
Claude Haiku 5.5Anthropic$0.201,425
GPT-4.1OpenAI$3.501,415
Solar Pro 4Upstage$0.521,388
GPT-4.1 miniOpenAI$0.701,383
GPT-4.1 nanoOpenAI$0.181,322

Price

Gemini 4 Carbon pricing

It is the #14 most expensive of 55 current text models priced per 1M input tokens.

UnitPrice (USD)TierAppliesSource
1M input tokens$2standardCurrent
1M output tokens$10standardCurrent
1M input tokens$0.10cachedCurrent

Benchmarks

Independent benchmark results

Scores come from independent evaluators that ran the model themselves, pinned to a dated snapshot. Rank is among current models in the same snapshot. Google's own launch figures are in the next section, because they are not measured the same way.

BenchmarkScoreRankMeasuredSource
Arena Text1528 Elo#1 of 52October 28, 2026Arena · 2026-11-03
Arena WebDev1768 Elo#4 of 46October 28, 2026Arena · 2026-11-03

Provider results

Reported by Google

Google's own results for Gemini 4 Carbon, beside the rivals it chose, as its table lists them. Labs pick the benchmarks and run them their own way, so compare within one table only. Internal benchmarks and other evaluators' indexes are left out.

Gemini 4 Carbon launch table, October 28, 2026Google · 2026-11-05

BenchmarkGemini 4 CarbonGemini 4 ArgonGPT-6 AstraClaude Opus 5.5Claude Fable 5.1
SEC-bench Pro88.2%83.4%85.4%Not listedNot listed
ExploitGym54.8%42.0%42.4%34.5%Not listed
CWE-bench71.0%68.0%68.0%67.0%58.0%
SWE-sweep9.6%4.3%9.5%3.6%2.9%
FrontierSWE 261.2%55.0%65.5%62.3%56.3%
Agents' Last Exam v1.1score77.4%69.0%73.1%71.6%Not listed
OSWorld-2.1offline subset81.5%74.7%80.9%87.2%76.8%
AutomationBenchZapier53.9%51.3%41.4%42.5%31.4%
Vending-Bench 2bank balance$16,484$13,718$15,515$9,235$5,422
Chartography72.5%71.6%71.0%66.3%46.2%
Blueprint-Bench 2similarity score58.6%54.4%49.7%51.2%41.9%
GraphWalks256K to 1M tokens, BFS F196.2%84.2%71.8%66.8%65.0%
RiemannBench80.0%76.0%72.0%69.6%65.6%

Google's table from the Gemini 4 Carbon announcement, evaluated without guardrails; Google publishes its methodology at deepmind.google/models/evals-methodology/gemini-4-carbon. The cybersecurity rows, Google's headline, come first. Left out here: the Artificial Analysis and Vals indexes and the two rows Artificial Analysis ran (Terminal-Bench 4.0 and Terminal-Bench-Science 0.1).

Fit

Strengths, cautions and best uses

Strengths
  • Google's strongest model at launch, built for cyber defense and long-horizon engineering: Google reports 88.2% on SEC-bench Pro for finding vulnerabilities, 54.8% on ExploitGym for exploitation and 71.0% on CWE-bench for patching, each ahead of GPT-6 Astra in its table
  • Arena rates it 1528 on Arena Text and 1768 on Arena WebDev (October 28, 2026), above Gemini 4 Argon on both
  • Google reports 96.2% on GraphWalks from 256K to 1M tokens, against 84.2% for Gemini 4 Argon, and an introductory price of $2 input and $10 output per 1M tokens, the same as Argon
Cautions
  • Limited access: Google gives it first to a small group of cyber defenders, then to more trusted partners through its Fairwind Program, which governments, healthcare providers and telecoms apply to; there is no public API or app access yet
  • The $2 and $10 rates are introductory, with cached input 95% off; after an introductory period Google has not dated, $4 input and $20 output per 1M tokens apply
  • In Google's own table it trails GPT-6 Astra on FrontierSWE 2 (61.2% against 65.5%) and Claude Opus 5.5 on OSWorld-2.1 (81.5% against 87.2%)
  • Google publishes no context window or output limit for Carbon yet
Best for
  • Finding, validating and patching vulnerabilities, for teams in Google's Fairwind Program
  • Long-horizon engineering over very long inputs, once access opens

From benchmark to production

Building on Gemini 4 Carbon?

Public scores rarely hold on your own data. We test it against the alternatives on your real tasks, then build the agent or pipeline around the model that wins.

AI Model Cost AuditLaunch price $100, regular price $150What the audit includes

Book the cost auditAI agent development

Alternatives

Alternatives to Gemini 4 Carbon

The closest current text models by price. The cheapest model that scores as well or better is listed first; failing that, the cheapest one within 3% of the benchmark's scale.

On score and price

Gemini 4 Carbon beside the 4 alternatives listed below, on Arena Text and on blended price per 1M tokens.

Gemini 4 Carbon and its alternatives on score and price.

Arena Text

Measured October 28, 2026, higher is better

Gemini 4 Carbon
1,528
Claude Sonnet 5
1,462
Claude Sonnet 5.5
1,477
GPT-4.1
1,415
GPT-5.6 Terra
1,467

Blended price per 1M tokens

Listed today, lower is better

Gemini 4 Carbon
$4
Claude Sonnet 5
$4
Claude Sonnet 5.5
$4
GPT-4.1
$3.50
GPT-5.6 Terra
$4.50

Further reading

More on Gemini 4 Carbon

Compare Gemini 4 Carbon with

All model comparisons

Articles about Gemini 4 Carbon

Other Gemini 4 models

FAQ

Gemini 4 Carbon FAQ

How much does Gemini 4 Carbon cost?

Gemini 4 Carbon costs $2 per 1M input tokens and $10 per 1M output tokens at the standard tier. It is the #14 most expensive of 55 current text models priced per 1M input tokens. Source: Google, verified November 5, 2026.

How does Gemini 4 Carbon score on independent benchmarks?

Gemini 4 Carbon scores 1528 Elo on Arena Text, #1 of 52 current models (Arena, measured October 28, 2026) and 1768 Elo on Arena WebDev, #4 of 46 current models (Arena, measured October 28, 2026). These are independent evaluations, not Google's own launch figures.

What benchmark results has Google published for Gemini 4 Carbon?

In Gemini 4 Carbon launch table (October 28, 2026), Google reports 88.2% on SEC-bench Pro, 54.8% on ExploitGym, 71.0% on CWE-bench and 9.6% on SWE-sweep. These are Google's own figures, beside the rivals it chose, not independent measurements.

Is Gemini 4 Carbon open source or API only?

Neither. Google offers Gemini 4 Carbon only to organizations it approves, with no public API and no published weights; the cautions on this page say who can apply.

What are the best alternatives to Gemini 4 Carbon?

The closest current alternatives by price and measured performance are Claude Sonnet 5 (same price per 1M input tokens, $2), Claude Sonnet 5.5 (same price per 1M input tokens, $2), GPT-4.1 (same price per 1M input tokens, $2) and GPT-5.6 Terra (same price per 1M input tokens, $2).

Sources

Sources

All 241 sources behind the benchmarks, with their versions, licenses and notes, are in the source catalog.

Cite this page

Van Data Team. "Gemini 4 Carbon Benchmarks, Pricing and Specs." AI Model Benchmark Dataset 2026-11-05. https://vandatateam.com/ai-model-benchmarks/gemini-4-carbon

Reuse the data: JSON · CSV · or explore every model in the AI model benchmark comparison.

Production evaluation

Deciding whether Gemini 4 Carbon fits your workload?

Public benchmarks narrow the shortlist. We test the finalists on your own tasks, data and cost limits, then build the routing and review controls to run the winner in production.

  • Evaluation cases and acceptance thresholds tied to real work
  • Model routing, fallbacks and cost controls
Compare all modelsBook a consultation