---
{
  "availableLangs": [
    "zh",
    "ar",
    "de",
    "en",
    "es",
    "fr",
    "hi",
    "id",
    "it",
    "ja",
    "ko",
    "ms",
    "nl",
    "pt",
    "ru",
    "th",
    "tr",
    "vi",
    "zh-hant"
  ],
  "categories": [
    "category-science-technology"
  ],
  "createdAt": "2026-10-30T14:52:31+08:00",
  "creators": [
    "Google DeepMind"
  ],
  "id": "work_gemini_4_carbon",
  "lang": "zh-hant",
  "published": {
    "day": 28,
    "display": "2026年10月28日",
    "month": 10,
    "precision": "day",
    "year": 2026
  },
  "revision": 20,
  "slug": "gemini-4-carbon",
  "sourceLang": "zh",
  "sources": [
    {
      "author": "Google",
      "date": "2026-10-28",
      "id": "s1",
      "title": "Gemini 4 Carbon: our next frontier model",
      "url": "https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/"
    },
    {
      "author": "9to5Google",
      "date": "2026-10-28",
      "id": "s2",
      "title": "Google announces Gemini 4 Carbon with ‘significant gains across coding and agentic tasks’",
      "url": "https://9to5google.com/2026/10/28/gemini-4-carbon-announcement/"
    },
    {
      "author": "Ars Technica",
      "date": "2026-10-28",
      "id": "s3",
      "title": "Google announces Gemini 4 Carbon, its latest frontier AI model that you can’t use",
      "url": "https://arstechnica.com/ai/2026/10/google-announces-gemini-4-carbon-its-latest-frontier-ai-model-that-you-cant-use"
    }
  ],
  "status": "published",
  "summary": "Google DeepMind 於 2026 年 10 月 28 日發布的前沿人工智慧模型，側重網路安全防禦與長週期任務，初期僅向受信任的網路安全人員開放。",
  "tags": [],
  "title": "Gemini 4 Carbon",
  "type": "work",
  "updatedAt": "2026-10-30T14:52:31+08:00",
  "updatedBy": "user:admin",
  "workKind": "ai_model"
}
---

# Gemini 4 Carbon

Google DeepMind 於 2026 年 10 月 28 日發布的前沿人工智慧模型，側重網路安全防禦與長週期任務，初期僅向受信任的網路安全人員開放。

Gemini 4 Carbon 是 [Google DeepMind](/zh-hant/organizations/google-deepmind.html) 於 2026 年 10 月 28 日發布的前沿人工智慧模型，[Google](/zh-hant/organizations/google.html) 稱其為迄今能力最強的模型，重點提升網路安全防禦與長週期任務處理能力。

## 發布與開放範圍

Carbon 與 9 月 30 日發布的 [Gemini 4 Argon](/zh-hant/works/gemini-4-argon.html) 沿用同一套安全與防護措施。發布時僅向一小批經篩選的網路安全防禦人員開放，隨後透過 Fairwind 計畫擴大至更多受信任合作夥伴，再面向開發者、企業和消費者推出。Google 表示正在參與美國政府的前沿模型發布前自願測試流程。

## 評測結果

Artificial Analysis Intelligence Index v5.0 給出 59 分，Carbon 在該指數中排名第一；WebDev Arena 得分 1768，排名第四。Google 公布的網路安全評測在未啟用防護欄的條件下進行：SEC-bench Pro 漏洞發現得分 88.2%，ExploitGym 漏洞可利用性驗證得分 54.8%，CWE-bench（Collinear）根因修復得分 71.0%。Google 同時稱，生成安全程式碼的成功率較 Argon 提升一倍以上，高嚴重度失配警報減少近 80%，Gray Swan 間接提示注入攻擊成功率為 0.0%。

## 定價

介紹期價格與 Argon 相同，為每百萬輸入 token 2 美元、輸出 token 10 美元，快取輸入按輸入價的 95% 折扣計費；介紹期結束後為每百萬輸入 token 4 美元、輸出 token 20 美元。

更新时间：2026-10-30

## 标签


## 关系

- 出版方：[Google DeepMind](/zh-hant/organizations/google-deepmind.html) 

第 1 / 1 页，共 1 条。

## 来源

- Gemini 4 Carbon: our next frontier model：https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/ 
- Google announces Gemini 4 Carbon with ‘significant gains across coding and agentic tasks’：https://9to5google.com/2026/10/28/gemini-4-carbon-announcement/ 
- Google announces Gemini 4 Carbon, its latest frontier AI model that you can’t use：https://arstechnica.com/ai/2026/10/google-announces-gemini-4-carbon-its-latest-frontier-ai-model-that-you-cant-use 
