Support Center
Frequently asked questions, quick-start guides, troubleshooting manual, and technical assistance.
Quick Resources & Guides
Essential developer guides and core platform shortcuts
Quick Start Guide
Learn Base URL configuration and OpenAI SDK / Python / cURL examples
Model Pricing Square
Explore all available models, context windows, and real-time pricing ratios
Quota & Top Up
Online top-up and redemption code activation. Quota never expires
工单支持
提交技术问题或业务咨询，在线与工程师交流
Frequently Asked Questions (FAQ)
Curated answers to the most common questions asked by developers
This platform is fully compatible with the standard OpenAI API specifications. In any OpenAI-supported SDK or client (such as NextChat, ChatBox, Cherry Studio, LangChain, Cursor, Dify), simply change the Base URL to:
* Note: Generate your secret key in "Console - API Keys" and use it as your Bearer Token / API Key.
HTTP Error Code Reference
Quickly identify and resolve issues when encountering API response errors
| Status | Error Type / Meaning | Common Causes | Troubleshooting & Solutions |
|---|---|---|---|
| 400 | Malformed request payload Bad Request | Malformed JSON, missing required parameters, non-existent model name, or prompt exceeded context limit. | Validate JSON format, verify model name against Model Square catalog, or reduce prompt length. |
| 401 | Invalid API Key / Unauthorized Unauthorized | Missing Authorization header, typo in API Key, or key expired/disabled. | Check token status in Console and ensure header is `Authorization: Bearer sk-xxxx`. |
| 403 | Access Forbidden Forbidden | Account banned, caller IP not in token whitelist, or key lacks permission for the requested model. | Inspect token IP and model restrictions in Console, or contact support for account review. |
| 429 | Insufficient Quota or Rate Limit Too Many Requests | Zero remaining balance, or request per minute (RPM) / token per minute (TPM) limits exceeded. | Top up quota in Console wallet; add retry delay with exponential backoff for high concurrency. |
| 500 / 502 | Gateway or Upstream Channel Error Internal / Upstream Error | Upstream provider outage or temporary network jitter. Gateway automatically fails over. | Platform auto-retries across redundant channels. Retry in 1-2s; check probes if persists. |
| 504 | Upstream Generation Timeout Gateway Timeout | Long generation prompt or heavy multimodal task exceeded HTTP socket timeout. | Enable `stream: true` for streaming output and increase your client HTTP timeout setting. |
Still Have Questions or Need Help?
Our engineering and support team is here to assist you
在线工单支持
提交工单描述您遇到的问题，可上传截图，工程师将尽快处理回复。
Official Email Support
For complex integration questions or enterprise inquiries, drop us a message