Are thread runs being throttled?

I now consistently see delays in the region of 5 seconds in thread run completion that were not apparent the day before yesterday.
Using gpt-4o, via V2 of assistants API.
Even simple prompts, e.g. “Hello”, result in the same observation.

Is anyone else able to reproduce?

Closing, non-issue.
I was inadvertently utilising the same thread with a history of 3700189 tokens :roll_eyes:

1 Like