{
  "contentStorage": "postgres",
  "title": "Gemini 3.1 Flash-Lite · संबंध",
  "description": "Google DeepMind 于 2026 年 3 月 3 日 发布的 Gemini 3.1 Flash-Lite。",
  "url": "/hi/works/gemini-3-1-flash-lite/relation.html",
  "entity": {
    "id": "work_gemini_3_1_flash_lite",
    "type": "work",
    "title": "Gemini 3.1 Flash-Lite",
    "slug": "gemini-3-1-flash-lite",
    "summary": "Google DeepMind 于 2026 年 3 月 3 日 发布的 Gemini 3.1 Flash-Lite。",
    "status": "published",
    "categories": [
      "category-science-technology"
    ],
    "tags": [],
    "sources": [
      {
        "id": "s1",
        "title": "Gemini 3.1 Flash Lite: Our most cost-effective AI model yet",
        "url": "https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-3-1-flash-lite/"
      },
      {
        "id": "s2",
        "title": "Gemini 3.1 Flash-Lite - Model Card — Google DeepMind",
        "url": "https://deepmind.google/models/model-cards/gemini-3-1-flash-lite/"
      },
      {
        "id": "s3",
        "title": "Gemini Developer API pricing",
        "url": "https://ai.google.dev/gemini-api/docs/pricing"
      },
      {
        "id": "s4",
        "title": "Model Evaluation – Approach, Methodology \u0026 Results: Gemini 3.1 Flash-Lite",
        "url": "https://deepmind.google/models/evals-methodology/gemini-3-1-flash-lite"
      }
    ],
    "createdAt": "2026-11-11T13:45:07+08:00",
    "updatedAt": "2026-11-11T13:45:07+08:00",
    "updatedBy": "user:admin",
    "revision": 1,
    "body": "## 发布背景\n\nGemini 3.1 Flash-Lite 是 [[organization_google_deepmind|Google DeepMind]] 于 2026 年 3 月 3 日 发布的 Gemini 模型[^s1]。2026年3月3日以预览版通过 Gemini API 与 Vertex AI 发布，是 Gemini 3 系列中最快、最具成本效益的模型。[^s1]\n\n## 定价与获取\n\n每百万 token 输入 0.25 美元、输出 1.50 美元（文本/图像/视频输入；音频输入为 0.50 美元）[^s3]。API 标识为 gemini-3.1-flash-lite[^s1]。\n\n## 性能表现\n\n以下均为 Google DeepMind 在发布公告与技术报告中公布的厂商自报成绩，尚未经独立复现：\n\n- GPQA Diamond：86.9%（厂商自报），高于 Gemini 2.5 Flash，为官方对照表中最高分[^s1]。\n- MMMU-Pro：76.8%（厂商自报），高于 Gemini 2.5 Flash 的 66.7%，为官方对照表中最高分[^s1]。\n- MMMLU：88.9%（厂商自报），高于 Gemini 2.5 Flash 的 86.6%，为官方对照表中最高分[^s1]。\n- Video-MMMU：84.8%（厂商自报），高于 Gemini 2.5 Flash 的 79.2%，为官方对照表中最高分[^s1]。\n\n## 主要特点\n\n上下文窗口 1,048,576 token，最大输出 65,536 token[^s1]。\n\n该模型基于 [[work_gemini_3_pro|Gemini 3 Pro]]，支持文本、图像、音频与视频输入，输出为文本，并提供可调的思考级别[^s1]。\n\n相比 [[work_gemini_2_5_flash|Gemini 2.5 Flash]]，首次回答时间快 2.5 倍、输出速度提升 45%（据 Artificial Analysis，厂商引用）[^s1]。\n\n## 定位\n\n它是 Gemini 3 系列的轻量高效档，接替 [[work_gemini_2_5_flash_lite|Gemini 2.5 Flash-Lite]]，基于 Gemini 3 Pro，后继为基于它的 Gemini 3.5 Flash-Lite。截至 2026 年 11 月，该模型仍在提供。",
    "url": "/works/gemini-3-1-flash-lite.html",
    "lang": "hi",
    "sourceLang": "zh",
    "fallback": true,
    "availableLangs": [
      "zh"
    ],
    "workKind": "ai_model"
  },
  "entities": null,
  "relations": [
    {
      "id": "relation_work_gemini_3_1_flash_lite_published_by_organization_google_deepmind",
      "fromId": "work_gemini_3_1_flash_lite",
      "relationType": "published_by",
      "toId": "organization_google_deepmind",
      "sourceRefs": [
        {
          "entityId": "work_gemini_3_1_flash_lite",
          "sourceId": "s1"
        }
      ],
      "label": "Published by",
      "targetId": "organization_google_deepmind",
      "targetTitle": "Google DeepMind",
      "targetUrl": "/hi/organizations/google-deepmind.html",
      "incoming": false
    }
  ],
  "mentions": [
    {
      "fromId": "work_gemini_3_5_flash_lite",
      "fromTitle": "Gemini 3.5 Flash-Lite",
      "fromUrl": "/hi/works/gemini-3-5-flash-lite.html",
      "fromType": "work"
    }
  ],
  "relationGroups": [
    {
      "type": "published_by",
      "label": "Published by",
      "items": [
        {
          "id": "relation_work_gemini_3_1_flash_lite_published_by_organization_google_deepmind",
          "fromId": "work_gemini_3_1_flash_lite",
          "relationType": "published_by",
          "toId": "organization_google_deepmind",
          "sourceRefs": [
            {
              "entityId": "work_gemini_3_1_flash_lite",
              "sourceId": "s1"
            }
          ],
          "label": "Published by",
          "targetId": "organization_google_deepmind",
          "targetTitle": "Google DeepMind",
          "targetUrl": "/hi/organizations/google-deepmind.html",
          "incoming": false
        }
      ]
    }
  ],
  "lang": "hi",
  "availableLangs": [
    "zh"
  ],
  "fallback": true,
  "infoFields": [
    {
      "key": "work_kind",
      "label": "Work kind",
      "value": "एआई मॉडल"
    }
  ],
  "pagination": {
    "page": 1,
    "pageSize": 50,
    "total": 1,
    "pageCount": 1
  },
  "relationsTotal": 1,
  "mentionsTotal": 1
}