{
 "generated": "2026-11-06",
 "sha": "78bbd39a8828",
 "node": {
  "id": "model:gemini-omni-1-2-flash",
  "type": "Model",
  "in_eps": [
   110
  ],
  "name": "Gemini Omni 1.2 Flash",
  "slug": "gemini-omni-1-2-flash",
  "date": "2026-11-05",
  "category": "video-world",
  "category_zh": "图像 · 视频 · 世界",
  "access": "closed",
  "eps": [
   110
  ],
  "page": "/graph/models/#m-gemini-omni-1-2-flash",
  "size": "视频生成与编辑模型，原生最长 15 秒（Omni 1.1 是 10 秒），3 到 15 秒可按秒设定；原生出 360p 或 720p，1080p 和 4K 靠上采样；Google 不披露参数规模",
  "avail": "闭源，Gemini API 和面向企业的 Gemini 产品里首次正式可用（GA）；官方写价格比 Omni 1.1 降了 30%，720p 每秒 $0.07；Google Vids 里有 Google 账号的用户每月可免费生成 3 条 15 秒视频；Gemini app 里先给 Google AI Ultra 订阅者",
  "license_full": "闭源，按 Google 的服务条款使用；输出带不可见的 SynthID 水印",
  "facts": [
   "官方写新增原生 15 秒生成、多轮对话式编辑、更强的参考图跟随；参考输入可以是角色、商品或风格图、不超过 3 秒的动作片段和音频提示（官方博客）",
   "官方博客也写：快速的体育动作、剧本快速切换和复杂物理还在继续改进",
   "Arena 官方帖说它在文生视频和图生视频榜上排第一、视频编辑榜第三（Arena 自己的帖子，分数以榜单页为准，本刊没有逐项核）"
  ],
  "sources": [
   {
    "label": "Google：Gemini Omni 1.2 Flash",
    "url": "https://blog.google/innovation-and-ai/technology/developers-tools/introducing-gemini-omni-1-2-flash/",
    "kind": "official"
   },
   {
    "label": "Google Vids：免费生成 15 秒视频",
    "url": "https://blog.google/products-and-platforms/products/workspace/omni-1-2-flash-google-vids/",
    "kind": "official"
   }
  ],
  "ep_tags": {
   "110": "官方博客 11-05"
  },
  "developer_text": "Google（Google DeepMind）",
  "degree": 0.0008
 },
 "page": "https://xiaopingfeng.com/graph/e/model/gemini-omni-1-2-flash/",
 "out": [
  {
   "type": "developed_by",
   "sentence": "Gemini Omni 1.2 Flash的开发方是Google",
   "direction": "out",
   "id": "org:google",
   "name": "Google",
   "unit": "Google DeepMind",
   "role": "primary",
   "status": "base",
   "evidence": [
    {
     "kind": "ledger",
     "label": "本站新模型台账",
     "text": "新模型台账：开发者栏写作“Google（Google DeepMind）”；Google：Gemini Omni 1.2 Flash",
     "url": "https://blog.google/innovation-and-ai/technology/developers-tools/introducing-gemini-omni-1-2-flash/"
    }
   ]
  },
  {
   "type": "placed_in",
   "sentence": "Gemini Omni 1.2 Flash出现在 Landscape 的「图像、视频与世界模型」",
   "direction": "out",
   "id": "layer:ai-models--5",
   "name": "图像、视频与世界模型",
   "status": "derived",
   "evidence": [
    {
     "kind": "own",
     "label": "本站页面",
     "text": "AI 新模型台账：Gemini Omni 1.2 Flash（图像、视频与世界模型）",
     "url": "https://xiaopingfeng.com/graph/models/#m-gemini-omni-1-2-flash"
    }
   ]
  }
 ],
 "in": []
}