AnLinkAI

Troubleshooting

Error handling for the current current integration path.

Use this page when a request fails and you need a fast decision tree. Start with request path, key state, model status, and account balance before assuming a deeper platform bug.

Errors

Common failure categories

The current API exposes a focused surface, so most failures still cluster into a small number of causes. Confirm model availability in the model catalog and balance steps in Billing and Recharge.

401 Unauthorized

The request is missing a valid bearer token or the API key is invalid.

  • Check Authorization header format
  • Confirm the key was copied correctly
  • Confirm the key has not been revoked

403 Forbidden

The credential is valid but cannot perform the requested action.

  • Check whether the route is admin-only
  • Check current account role in the console
  • Retry with the expected account

404 Not Found

The request path is wrong or the model code is not exposed by the current route.

  • Confirm you are calling /chat/completions
  • Confirm the base URL is correct
  • Confirm the model code matches the exposed model catalog exactly

429 / Balance rejection

The request cannot proceed because balance or limits do not permit it.

  • Check remaining USDT balance in the console
  • Reduce request size for validation
  • Check whether trial balance has been exhausted

Model disabled

The model exists in the catalog but is not currently active.

  • Open the model catalog
  • Confirm the status field is active
  • Switch to another active model for test traffic

Upstream request failure

The request reached the platform but failed during upstream execution or routing.

  • Capture the request ID
  • Check Usage Logs in the console
  • Retry with a smaller prompt and simpler payload

UPSTREAM_STREAM_INTERRUPTED

A streaming request started but the upstream SSE connection ended unexpectedly.

  • Retry once with a small prompt
  • Check Usage Logs for the matching stream request
  • Use non-streaming mode if the provider route is unstable

STREAM_USAGE_MISSING

The stream completed but the upstream did not return usage metadata, so AnLinkAI did not deduct cost for that call.

  • Confirm stream_options.include_usage is enabled
  • Check whether the selected upstream supports stream usage
  • Retry with qwen-flash or deepseek-v4-flash

CLIENT_STREAM_CLOSED

The client closed the streaming connection before the request completed.

  • Check client timeout settings
  • Avoid closing the browser tab or HTTP connection early
  • Retry with a smaller max_tokens value

Invalid request body

The JSON shape or required fields are malformed.

  • Confirm model is present
  • Confirm messages is a JSON array
  • Confirm Content-Type is application/json

Fast Path

Minimal troubleshooting order

  1. Check the request path and base URL first.
  2. Check whether the API key is still active.
  3. Check whether the selected model is marked active in the model catalog.
  4. Check remaining balance in the console.
  5. Check Usage Logs for the matching request entry.

Console URL

https://console.anlinkai.com

Primary path

https://api.anlinkai.com/api/v1/chat/completions

Next page

/faq

Model catalog

/model-catalog

Billing

/billing-and-recharge