Gemini 4 Carbon logo

Gemini 4 Carbon

Google · Sorti oct. 2026

Conditionnel

Conditional. Gemini 4 Carbon is Google's strongest model by its own account, announced October 28, 2026 with what Google calls state-of-the-art cyber results, but access is limited to a subset of cyber defenders and trusted partners through the Fairwind Program. Every figure is Google's own and we have not tested it. Right for security teams that can qualify; everyone else has to wait for the broader release, which Google has not dated.

Est-ce fait pour vous ?

Recommandé pour

  • Vulnerability discovery, exploitation and patching research; Google reports SEC-bench Pro 88.2%, ExploitGym 54.8% and CWE-bench 71.0% (Google's claims, untested by us)
  • Secure code generation; Google says Carbon delivers more than 2x uplift in success rate over Argon (Google's claim)
  • Security agents exposed to untrusted content; Google reports 0.0% attack success on Gray Swan's indirect prompt injection benchmark, down from Argon's 0.7% (Google's claim)
  • Sustained agentic coding and long-horizon work through the Gemini API, where Google lists gemini-4-carbon as generally available with a 1,048,576-token context window

Déconseillé pour

  • Budgeting past November 30, 2026: the introductory $2 input and $10 output per 1M tokens ends then, and the rate becomes $4 and $20
  • Free-tier use; Google's pricing page says Carbon is not available on the Free tier and every token is billed
  • Assuming cyber capabilities are unrestricted; the launch post limited them to Fairwind defenders and the model page we read does not repeat or lift that limit, so confirm with Google
  • Decisions that need independent benchmark confirmation; every figure comes from Google's announcement, including the Artificial Analysis figure of 59, which we did not check against Artificial Analysis

Performances par tâche

Vulnerability discovery

Very Good

SEC-bench Pro 88.2%, called state-of-the-art by Google; vendor claim, untested by us

Exploit generation

Very Good

ExploitGym 54.8%, called state-of-the-art by Google; vendor claim, untested by us

Patching

Very Good

CWE-bench 71.0% on Collinear's benchmark per Google; vendor claim, untested by us

Prompt-injection robustness

Very Good

0.0% attack success on Gray Swan's indirect prompt injection benchmark per Google, against Argon's 0.7%; vendor claim, untested by us

Tarifs

Entrée

$2 / 1M tokens (intro)

Sortie

$10 / 1M tokens (intro)

Contexte

1M ctx, 64K out; $2/$10 to Nov 30, then $4/$20

Voir tous les tarifs

Tests de performance

TestScoreSource
SEC-bench Pro (Google-reported)88.2% Source
ExploitGym (Google-reported)54.8% Source
CWE-bench (Google-reported)71.0% Source
WebDev Arena (Google-reported)#4 (1768) Source
Artificial Analysis Intelligence Index v5.0 (as stated by Google)59 (#1) Source
Gray Swan indirect prompt injection attack success rate (Google-reported, lower is better)0.0% Source

Aucun changement de verdict pour l’instant

L’horloge tourne dès le premier jour : les changements apparaîtront ici à mesure que notre verdict évolue.

Journal de vérification

  • Statut· Mis à jour

    Agent automatisé

    Now generally available on the Gemini API as gemini-4-carbon (changelog Nov 10, 2026), 1,048,576 context, 65,536 default output, intro $2/$10 through Nov 30 then $4/$20. Stored entity said gated, no API id. Refreshed pricing_context, strengths, audiences, sources; verdict summary restage owed.

  • Tarifs· Aucun changement

    Agent automatisé

Comment nous évaluons