API GPU acquisition failure — perpetually stuck at "Waiting for a GPU to become available

Beginners
1

I am currently experiencing a persistent GPU acquisition failure when calling the API. My requests get stuck indefinitely on the message: "Waiting for a GPU to become available".

Current behavior Despite waiting, no GPU is ever allocated and the request never progresses past the waiting state. I have been facing this exact issue consistently for the past 2 days across all API attempts.

Is anyone else experiencing this, or is there a known limitation/outage with the API GPU queues right now? Any help to unblock this would be appreciated.

2

P.S.
Current behavior Despite waiting for well over 60 seconds, no GPU is ever allocated and the request never progresses past the waiting state.

To clarify, this does not appear to be a standard quota limitation issue. Normally, when I exceed my quota, the system returns a clear error immediately, such as: gradio.exceptions.Error: 'You have exceeded your ZeroGPU quota (78s requested vs. 47s left). Try again in 13:33:02. Authenticate with a Hugging Face token for more quota'

3

P.S. I am a paid user and I am fully authenticated/logged in when making these API calls. However, when the request finally fails after 60 seconds, the exact traceback I receive is: spaces.zero.gradio.HTMLError: No GPU was available after 60s [Create a free account](https://huggingface.co/join) to get a higher priority in ZeroGPU queues.

This error message is highly confusing because it treats my request as if I am an unauthenticated guest user, completely ignoring my logged-in and paid tier status. It seems the queue is not recognizing my token/priority correctly.