API Docs

Get started in three steps — compatible with every OpenAI client.

Step 1: note the Base URL

This is the root address for all your requests. Use it as the API base in your client.

Base URL https://ltzy.zhuafs.com/v1

Use exactly this address. Do not modify it and do not append an extra /v1. Most connection errors come from a wrong Base URL.

Step 2: create an API key

Create a key

Step 3: make your first call

Replace sk-your-key below with your key and run it in a terminal.

Step 3: make your first call

curl https://ltzy.zhuafs.com/v1/chat/completions \
  -H "Authorization: Bearer sk-your-key" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "deepseek-v4.1-flash",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

Popular client setup

Client	Key settings
ChatGPT-Next-Web	Settings → Custom Endpoint: fill the Base URL above; API Key: your key.
LobeChat	Settings → Language Model → OpenAI: set API Proxy Address to the Base URL.
Cherry Studio	Model Service → Add OpenAI provider: API Host = Base URL, then add the key.
Cursor	Settings → Models → OpenAI API Key + Override Base URL: fill the Base URL.
Cline (VS Code)	API Provider: OpenAI Compatible; Base URL: the Base URL above; enter the Model ID manually.
Dify	Model Provider → OpenAI: set API Base (without /v1 suffix handling issues) and the key.
Open WebUI	Settings → Connections → OpenAI API: Base URL = the Base URL above.
Immersion Translate	AI service → OpenAI: custom interface address = Base URL, fill key and model.

Error troubleshooting

Code	Cause & fix
`401`	Missing or invalid API key. Check that the key is complete and not deleted.
`402`	Insufficient balance or key quota. Top up or use a key without a quota limit.
`403`	Model not allowed for this key, or the account is disabled.
`404`	Wrong URL. Make sure it ends with /v1/chat/completions.
`413`	Request too large. Reduce the input size.
`429`	Rate limited or upstream busy. Retry after a short wait.
`500 / 502 / 503`	Upstream or gateway issue. Retry; if it persists, check the Status page.
`Timeout`	The upstream is slow. Try a smaller request or another model.