After deploying a service on Elastic Algorithm Service (EAS), use the online debugging feature in the PAI console to send test requests and inspect responses without writing any code.
Not all services support online debugging. For ComfyUI and SDWebUI, only the Standard Edition supports this feature. If your service type does not support online debugging, use the service endpoint directly with a REST client. See Service invocation methods for details.
Prerequisites
Before you begin, make sure that you have:
A service deployed on EAS
Access to the PAI console
Send a test request
Log on to the PAI console and select a region.
Select a workspace, then click Elastic Algorithm Service (EAS).
On the Inference Service tab, find your service and click
> Online Debugging in the Actions column.On the online debugging page, configure the following request parameters:
Parameter Description Method The HTTP method. Usually POSTorGET.URI of the request (URL) The base URL is auto-populated. Append the specific API operation path required by your model (for example, /v1/chat/completions). An incorrect path returns a404 Not Founderror.Request header (Headers) The Authorizationtoken is auto-populated. No changes are needed in most cases.Request body (Body) The request payload. Must match the format required by your model's API. An incorrect format returns a 400 Bad Requestor500 Internal Server Errorerror.ImportantThe URL path and request body format depend on the model you deployed. Refer to the model's API documentation to construct them correctly.
Send the request. A successful response displays the model output. If the request fails, check the HTTP status code and error message in the response — the most common causes are an incorrect URL path or a malformed request body.
Example: test a vLLM-deployed model
The following example tests the chat completions API for a Qwen2.5-7B-Instruct model service deployed with vLLM.
Request configuration:
| Parameter | Value |
|---|---|
| Method | POST |
| URI of the request (URL) | http://***********/v1/chat/completions (append /v1/chat/completions to the auto-populated base URL) |
Request body:
{
"model": "Qwen2.5-7B-Instruct",
"messages": [
{
"role": "user",
"content": "What is the capital of Canada?"
}
]
}Result:
FAQ
How do I find the correct URL path and request body format?
The correct URL path and request body depend on the model you deployed. See Construct requests for typical scenarios for model-specific examples.
Next steps
Stress test your service: Use automatic service stress testing to evaluate performance under load.
Invoke in production: After the service meets your requirements, see Service invocation methods to integrate it into your application.