向 OpenAI API 發出請求

Working with the OpenAI API

James Chapman

Curriculum Manager, DataCamp

API 端點

電腦透過多個存取點連到 OpenAI，這些存取點再連到不同伺服器。

API 端點

電腦透過多個端點連到 OpenAI，端點再連到不同伺服器。

API 端點

端點（Endpoints） → 為特定互動設計的 API 存取點

醫院走廊，通往不同科別的多扇門。

API 驗證

驗證（Authentication） → 控制對 API 端點服務的存取（通常使用唯一的 key）

醫院走廊，通往不同科別的多扇門。

API 使用成本

針對 OpenAI API：
- 模型
- 輸入與輸出越大 → 成本越高

本課程不需額外費用 🎉

收費高速公路。

¹ https://openai.com/pricing

建立 API 金鑰

本課程不需要

建立帳號 → https://platform.openai.com/signup
前往 API keys 頁面 → https://platform.openai.com/account/api-keys
建立新的 secret key 並複製

建立新 secret key 的按鈕。

發出請求

from openai import OpenAI


client = OpenAI(api_key="ENTER YOUR KEY HERE")


response = client.chat.completions.create(

    model="gpt-4o-mini",

    messages=[{"role": "user", "content": "What is the OpenAI API?"}]

)


print(response)

回應內容

ChatCompletion(id='chatcmpl-AEcQbQekIzxcxVAKYVAjgUAXokgrl', choices=[Choice(finish_reason='length', index=0,
logprobs=None, message=ChatCompletionMessage(content='The OpenAI API is a cloud-based service provided by
OpenAI that allows developers to integrate advanced AI models into their applications.', refusal=None,
role='assistant', function_call=None, tool_calls=None))], created=1728047673, model='gpt-4o-mini-2024-07-18',
object='chat.completion', service_tier=None, system_fingerprint='fp_f85bea6784',
usage=CompletionUsage(completion_tokens=30, prompt_tokens=14, total_tokens=44,
prompt_tokens_details={'cached_tokens': 0}, completion_tokens_details={'reasoning_tokens': 0}))

回應內容

ChatCompletion(id='chatcmpl-AEcQbQekIzxcxVAKYVAjgUAXokgrl',
               choices=[Choice(finish_reason='length', index=0, logprobs=None,
                               message=ChatCompletionMessage(content='The OpenAI API is a cloud-based service
                               provided by OpenAI that allows developers to integrate advanced AI models
                               into their applications.',
                               refusal=None, role='assistant', function_call=None, tool_calls=None))],
               created=1728047673,
               model='gpt-4o-mini-2024-07-18',
               object='chat.completion', service_tier=None, system_fingerprint='fp_f85bea6784',
               usage=CompletionUsage(completion_tokens=25, prompt_tokens=14, total_tokens=39,
                                     prompt_tokens_details={'cached_tokens': 0},
                                     completion_tokens_details={'reasoning_tokens': 0}))

解讀回應

print(response.choices)

[Choice(finish_reason='length', index=0, logprobs=None,
 message=ChatCompletionMessage(content='The OpenAI API is a cloud-based service provided by
 OpenAI that allows developers to integrate advanced AI models into their applications.',
 refusal=None, role='assistant', function_call=None, tool_calls=None))]

解讀回應

print(response.choices[0])

Choice(finish_reason='length', index=0, logprobs=None,
 message=ChatCompletionMessage(content='The OpenAI API is a cloud-based service provided by
 OpenAI that allows developers to integrate advanced AI models into their applications.',
 refusal=None, role='assistant', function_call=None, tool_calls=None))

解讀回應

print(response.choices[0].message)

ChatCompletionMessage(content='The OpenAI API is a cloud-based service provided by
 OpenAI that allows developers to integrate advanced AI models into their applications.',
 refusal=None, role='assistant', function_call=None, tool_calls=None)

print(response.choices[0].message.content)

The OpenAI API is a cloud-based service provided by OpenAI that allows developers to
integrate advanced AI models into their applications.

一起來練習吧！

Working with the OpenAI API