GPT-4o-audio

POSThttps://www.aipro2.com/v1/chat/completions

官方文档：https://platform.openai.com/docs/api-reference/chat/create

请求体参数

application/json

• audioobject
  音频输出参数。
• messages*array<object>
  构成当前对话的消息列表。
• modalitiesarray<string>
  模型生成的输出类型。默认设置：["text"]。要请求此模型生成文本和音频响应，您可以使用：["text", "audio"]。
• model*string
  要使用的模型 ID。

响应

application/jsonV1ChatCompletionsResponse

• choices*array<object>
  候选结果列表
• created*integer
  创建时间（Unix 时间戳）
• id*string
  资源或任务 ID
• model*string
  模型名称 / 型号 ID
• object*string
  对象类型标识，如 chat.completion
• service_tier*string
  服务档位 / 优先级
• system_fingerprint*string
  后端配置指纹
• usage*object
  Token 用量统计

请求体示例示例 1

1{2  "audio": {3    "format": "wav",4    "voice": "alloy"5  },6  "messages": [7    {8      "content": "Is a golden retriever a good family dog?",9      "role": "user"10    }11  ],12  "modalities": [13    "text",14    "audio"15  ],16  "model": "gpt-4o-audio-preview"17}

代码示例cURLPythonJavaScriptGoPHP

1curl -X POST "https://www.aipro2.com/v1/chat/completions" \2  -H "Authorization: Bearer <API Key>" \3  -H "Content-Type: application/json" \4  -d '{5    "audio": {6      "format": "wav",7      "voice": "alloy"8    },9    "messages": [10      {11        "content": "Is a golden retriever a good family dog?",12        "role": "user"13      }14    ],15    "modalities": [16      "text",17      "audio"18    ],19    "model": "gpt-4o-audio-preview"20  }'

响应示例 · 200example

1{2  "choices": [3    {4      "finish_reason": "stop",5      "index": 0,6      "logprobs": "",7      "message": {8        "annotations": [],9        "content": "\n\nHello there, how may I assist you today?",10        "refusal": "",11        "role": "assistant"12      }13    }14  ],15  "created": 1677652288,16  "id": "chatcmpl-123",17  "model": "gpt-4o",18  "object": "chat.completion",19  "service_tier": "default",20  "system_fingerprint": "fp_abc123",21  "usage": {22    "completion_tokens": 12,23    "completion_tokens_details": {24      "accepted_prediction_tokens": 0,25      "audio_tokens": 0,26      "reasoning_tokens": 0,27      "rejected_prediction_tokens": 028    },29    "latency_checkpoint": {30      "engine_tbt_ms": 0,31      "engine_ttft_ms": 0,32      "engine_ttlt_ms": 0,33      "pre_inference_ms": 0,34      "service_tbt_ms": 0,35      "service_ttft_ms": 0,36      "service_ttlt_ms": 0,37      "total_duration_ms": 0,38      "user_visible_ttft_ms": 039    },40    "prompt_tokens": 9,41    "prompt_tokens_details": {42      "audio_tokens": 0,43      "cached_tokens": 044    },45    "total_tokens": 2146  }47}