错误处理与重试机制

Tokenlio 返回标准的 HTTP 响应状态码与格式一致的 JSON 错误负载。

错误结构

{
  "error": {
    "message": "Insufficient balance. Please top up your workspace wallet.",
    "type": "insufficient_balance",
    "param": null,
    "code": 402
  }
}

常见 HTTP 状态码

状态码	含义	说明与处理建议
200	OK	请求成功
400	Bad Request	请求参数格式错误或缺失必填项
401	Unauthorized	API Key 缺失、失效或格式错误
402	Payment Required	账户余额不足，请登录控制台充值
404	Not Found	请求的模型 ID 不存在或已下线
429	Too Many Requests	触发并发频率限制，网关支持自动加权旁路平滑切流
500/502	Server / Gateway Error	上游服务异常，建议实施指数退避重试

推荐的重试模式 (Exponential Backoff)

import time
from openai import OpenAI, RateLimitError, APIError

def call_with_retry(client, **kwargs):
    max_retries = 3
    for attempt in range(max_retries):
        try:
            return client.chat.completions.create(**kwargs)
        except (RateLimitError, APIError) as e:
            if attempt == max_retries - 1:
                raise e
            wait = 2 ** attempt
            time.sleep(wait)