curl -X POST "https://api.tokenops.ai/v1beta/models/gemini-2.5-flash:streamGenerateContent" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer <API-KEY>" \
  -d '{
    "contents": [
      {
        "parts": [
          {
            "text": "讲个300个字的故事"
          }
        ]
      }
    ],
    "generationConfig": {
      "maxOutputTokens": 3000,
      "temperature": 0.7
    }
  }' \
  --no-buffer

import requests
import json

# 配置API密钥和基础URL
API_KEY = "<API-KEY>"
BASE_URL = "https://api.tokenops.ai/v1beta"

def stream_chat_with_gemini(message, model="gemini-2.5-flash"):
    url = f"{BASE_URL}/models/{model}:streamGenerateContent"
    headers = {
        "Content-Type": "application/json",
        "Authorization": f"Bearer {API_KEY}"
    }
    
    data = {
        "contents": [
            {
                "parts": [
                    {
                        "text": message
                    }
                ]
            }
        ],
        "generationConfig": {
            "maxOutputTokens": 3000,
            "temperature": 0.7
        }
    }
    
    response = requests.post(url, headers=headers, json=data, stream=True)
    response.encoding = 'utf-8'  # 设置响应编码为UTF-8

    if response.status_code == 200:
        full_response = ""

        for line in response.iter_lines(decode_unicode=True):
            if line.strip():
                # 去掉 "data: " 前缀（如果有）
                if line.startswith("data: "):
                    line = line[6:]
                try:
                    data = json.loads(line)
                    if 'candidates' in data and len(data['candidates']) > 0:
                        candidate = data['candidates'][0]
                        if 'content' in candidate and 'parts' in candidate['content']:
                            for part in candidate['content']['parts']:
                                if 'text' in part:
                                    content = part['text']
                                    full_response += content
                                    print(content, end="", flush=True)
                except json.JSONDecodeError:
                    continue
                    
        return full_response
    else:
        return f"错误: {response.status_code} - {response.text}"

# 使用示例
if __name__ == "__main__":
    message = "讲个300个字的故事"
    print("Gemini回复: ", end="")
    reply = stream_chat_with_gemini(message)
    print(f"\n\n完整回复: {reply}")