Deepgram

Deepgram 是語音轉文字 API。在 OpenClaw 中，它透過 tools.media.audio 用於傳入音訊/語音訊息轉錄，並透過 plugins.entries.voice-call.config.streaming 用於 Voice Call 串流 STT。

對於批次轉錄，OpenClaw 會將完整音訊檔案上傳到 Deepgram，並將轉錄稿注入回覆管線（{{Transcript}} + [Audio] 區塊）。對於 Voice Call 串流，OpenClaw 會透過 Deepgram 的 WebSocket listen 端點轉送即時 G.711 u-law 影格，並在 Deepgram 傳回時發出部分或最終轉錄稿。

# 開始使用

DEEPGRAM_API_KEY=dg_...

{
  tools: {
    media: {
      audio: {
        enabled: true,
        models: [{ provider: "deepgram", model: "nova-3" }],
      },
    },
  },
}

# 設定選項

{
  tools: {
    media: {
      audio: {
        enabled: true,
        models: [{ provider: "deepgram", model: "nova-3", language: "en" }],
      },
    },
  },
}

{
  tools: {
    media: {
      audio: {
        enabled: true,
        providerOptions: {
          deepgram: {
            detect_language: true,
            punctuate: true,
            smart_format: true,
          },
        },
        models: [{ provider: "deepgram", model: "nova-3" }],
      },
    },
  },
}

# Voice Call 串流 STT

隨附的 deepgram Plugin 也會為 Voice Call Plugin 註冊即時轉錄供應商。

{
  plugins: {
    entries: {
      "voice-call": {
        config: {
          streaming: {
            enabled: true,
            provider: "deepgram",
            providers: {
              deepgram: {
                apiKey: "${DEEPGRAM_API_KEY}",
                model: "nova-3",
                endpointingMs: 800,
                language: "en-US",
              },
            },
          },
        },
      },
    },
  },
}

# 備註

# 相關