メインコンテンツまでスキップ

Prompt with stream

POST 
/bv/aisk-stream/v1/prompts

This API is used to prompt streaming with the chatbot. The authorization header should be prompt token.

It'll return a 429/Too Many Requests and consider the following error reasons:

  • With ERROR_REASON_PROMPT_THROTTLING that means the rate limit is reached and will temporarily throttle the chatbot which should reduce the requested rate to satisfy the chatbot rate limit.

    Response examples
    \{
      "code": 8,
      "message": "throttle",
      "details": [
        \{
          "@type": "type.googleapis.com/google.rpc.ErrorInfo",
          "reason": "ERROR_REASON_PROMPT_THROTTLING",
          "domain": "bv.aisk.v1",
          "metadata": \{
            "duration": "1m0s",
            "limit": "5",
            "throttle_message": "Too many messages in a row",
            "vary_by": "RATE_LIMIT_VARY_BY_DEVICE_ID",
            "wait": "36.581786s",
            "wait_sec": "36.581786"
          \}
        \}
      ]
    \}

  • With ERROR_REASON_PROMPT_OUT_OF_QUOTA that means the daily quota is reached and will be reset at 00:00 in the organization timezone.

    Response examples
    \{
      "code": 8,
      "message": "out of quota",
      "details": [
        \{
          "@type": "type.googleapis.com/google.rpc.ErrorInfo",
          "reason": "ERROR_REASON_PROMPT_OUT_OF_QUOTA",
          "domain": "bv.aisk.v1",
          "metadata": \{
            "limit": "100",
            "reset_at": "2024-01-06T00:00:00+08:00",
            "reset_in": "11h18m33.091627s",
            "reset_in_sec": "40713.091627"
          \}
        \}
      ]
    \}

Request​

Responses​

A successful response.