All models failed (4): zhipu/glm-4.7: ⚠️ API rate limit reached. Please try agai...

报错分析

1.错误翻译

原文：FallbackSummaryError: All models failed (4): zhipu/glm-4.7: ⚠️ API rate limit reached. Please try again later. (rate_limit) | zhipu/GLM-5-Turbo: ⚠️ API rate limit reached. Please try again later. (rate_limit) | minimax/MiniMax-M2.7: HTTP 401 authentication_error: login fail: Please carry the API secret key in the 'Authorization' field of the request header (auth) | minimax/MiniMax-M2.5: HTTP 401 authentication_error: login fail: Please carry the API secret key in the 'Authorization' field of the request header (auth)
译文：回退汇总错误：所有模型均失败（4个）：智谱glm-4.7：⚠️ 达到API速率限制，请稍后重试。(rate_limit) | 智谱GLM-5-Turbo：⚠️ 达到API速率限制，请稍后重试。(rate_limit) | Minimax/MiniMax-M2.7：HTTP 401认证错误：登录失败：请在请求头的'Authorization'字段中携带API密钥(auth) | Minimax/MiniMax-M2.5：HTTP 401认证错误：登录失败：请在请求头的'Authorization'字段中携带API密钥(auth)

2.原因分析

3. 解决方案

import requests
from tenacity import retry, wait_exponential, stop_after_attempt

# 1. 修复认证错误：统一设置Authorization头（从环境变量/配置文件取密钥）
API_KEYS = {
    "zhipu": "YOUR_ZHIPU_API_KEY",  # 替换为实际密钥
    "minimax": "YOUR_MINIMAX_API_KEY"  # 替换为实际密钥
}

# 2. 修复速率限制：添加指数退避重试
@retry(wait=wait_exponential(min=4, max=10), stop=stop_after_attempt(3))
def call_model(model_name, url, payload):
    headers = {"Content-Type": "application/json"}
    # 动态设置认证头
    if model_name.startswith("zhipu/"):
        headers["Authorization"] = f"Bearer {API_KEYS['zhipu']}"
    elif model_name.startswith("minimax/"):
        headers["Authorization"] = f"Bearer {API_KEYS['minimax']}"
    
    resp = requests.post(url, json=payload, headers=headers, timeout=30)
    resp.raise_for_status()  # 抛出HTTP错误（如401/429）
    return resp.json()

# 示例调用（替换为实际URL和参数）
try:
    result = call_model(
        model_name="zhipu/glm-4.7",
        url="https://open.bigmodel.cn/api/paas/v4/chat/completions",
        payload={"messages": [{"role": "user", "content": "Hello"}]}
    )
except Exception as e:
    print(f"调用失败：{str(e)}")

4. 预防措施