Openai API 기본적인 호출방식

 

 

RAG 를 사용하지 않은 기본적인 API Call 방법

 

1. 기본 Openai API

from openai import OpenAI
import os
os.environ["OPENAI_API_KEY"] = "sk..."

client = OpenAI()
completion = client.chat.completions.create(
    model="gpt-4o",
    temperature=0.1,
    messages=[
        {"role":"system", "content":system_prompt},
        {"role":"user", "content":user_query}
    ],
    response_format={"type":"json_object"}
)
response = completion.choices[0].message.content
return(response)

2. Llama_index 활용

from llama_index.llms.openai import OpenAI
from llama_index.core.llms import ChatMessage
import os
os.environ["OPENAI_API_KEY"] = "sk..."

# OpenAI 모델 설정
llm = OpenAI(model="gpt-4o", temperature=0.1)

messages = [
    ChatMessage(role="system", content=system_prompt),
    ChatMessage(role="user", content="Hi"),
]

# Response 를 json 형식으로 받고 싶은 경우 
resp = llm.chat(messages, response_format={"type": "json_object"})