Python-openai格式-流式

import requests
import json
import sys
import time

# API 配置
API_URL = "https://xinyun.ai/v1/chat/completions"
API_KEY = "sk-xxxxxxx"

def stream_chat_completion(query):
    headers = {
        "Content-Type": "application/json; charset=utf-8",
        "Authorization": f"Bearer {API_KEY}"
    }
    data = {
        "model": "gemini-2.0-flash",
        "messages": [{"role": "user", "content": query}],
        "stream": True
    }

    try:
        # 发送请求
        response = requests.post(API_URL, headers=headers, json=data, stream=True, timeout=30)
        response.encoding = 'utf-8'  # 强制指定 UTF-8 解码

        print("模型回答: ", end="", flush=True)
        for line in response.iter_lines():
            if line:
                # 直接解码为 UTF-8，避免重复解码
                line_str = line.decode('utf-8', errors='replace').strip()
                if line_str.startswith("data: "):
                    json_str = line_str[6:]
                    if json_str == "[DONE]":
                        break
                    try:
                        chunk = json.loads(json_str)
                        if "choices" in chunk and len(chunk["choices"]) > 0:
                            delta = chunk["choices"][0].get("delta", {})
                            content = delta.get("content", "")
                            if content:
                                # 逐字打印
                                for char in content:
                                    print(char, end="", flush=True)
                                    time.sleep(0.05)
                    except json.JSONDecodeError:
                        continue
        print()

    except requests.exceptions.RequestException as e:
        print(f"请求出错: {e}")

def main():
    # 配置标准输出为 UTF-8
    sys.stdout.reconfigure(encoding='utf-8')

    while True:
        query = input("\n请输入你的问题（输入 '退出' 结束）：")
        if query.strip().lower() == "退出":
            print("程序结束")
            break
        stream_chat_completion(query)

if __name__ == "__main__":
    main()