{
  "schemaVersion": "1.0",
  "updates": [],
  "companyNews": [],
  "id": "b882",
  "language": "zh",
  "name": "faster-whisper",
  "originalName": "faster-whisper",
  "library": "tools",
  "category": "语音识别与输入",
  "categoryKey": "语音识别与输入",
  "categoryKeys": [
    "语音识别与输入"
  ],
  "categories": [
    "语音识别与输入"
  ],
  "entryPoints": [],
  "summary": "基于 CTranslate2 的 Whisper 推理实现，支持 CPU/GPU 量化、批量转写、词级时间戳与语音活动过滤。",
  "introduction": "faster-whisper 通过 CTranslate2 运行兼容的 Whisper 模型，将音频转成文本及带时间信息的片段。Python 接口支持批量转写、词级时间戳、Silero VAD 过滤和语言检测，CPU 与 GPU 均可选择量化计算方式。适合把转写嵌入自建服务或离线批处理；运行前需准备模型，GPU 加速还依赖匹配的 CUDA、cuDNN 与 CTranslate2 版本，实际速度随模型和硬件而变。",
  "capabilities": [],
  "company": {
    "name": "SYSTRAN",
    "originalName": "SYSTRAN"
  },
  "officialUrl": "https://github.com/SYSTRAN/faster-whisper",
  "profileUpdatedOn": "2026-10-22",
  "sources": [
    {
      "kind": "product",
      "url": "https://github.com/SYSTRAN/faster-whisper"
    }
  ],
  "news": [],
  "links": {
    "interactive": "products/faster-whisper--b882.html",
    "reading": "read/zh/products/faster-whisper--b882.html",
    "markdown": "read/zh/products/faster-whisper--b882.md",
    "json": "read/zh/products/faster-whisper--b882.json"
  }
}