Skip to main content
The AI provider is temporarily overloaded. Wait a moment, then try again.

​
What happened

The model provider could not process your request because its service was overloaded or temporarily unavailable.

​
What you can do

  1. Wait a minute, then press Retry.
  2. If the error keeps returning, choose another available model and send the request again.
  3. If every model fails, contact Pretty support with the time of the error and the message ID.
Provider overloads usually clear without a settings change. Repeated failures across several minutes may point to a provider incident or a problem with your workspace’s connected AI service.

​
For admins

Check the provider’s status page first. If your workspace uses its own provider or LiteLLM proxy, also check the proxy logs and upstream provider response. Pretty shows this message for temporary provider errors such as server_error, server_is_overloaded, and vector_store_timeout. Direct provider connections also use it for HTTP 5xx responses. A LiteLLM gateway response with HTTP 502, 503, or 504 is treated as a proxy outage instead. If the proxy itself cannot be reached, Pretty shows a different message. See AI service isn’t available.