Hi,
I am making an API call to gpt-4.1-nano-2025-04-14 chat completion API. I pass it a prompt which is around 290 tokens and the output is around 600 tokens. This is taking around 10.5 seconds in bubble.
Is it expected to take this long for Open AI API calls in bubble?
The model has a latency of 0.44 seconds at the moment, and 600 output tokens at 80 tokens per second = 7.5 seconds.
So, 10 seconds is probably about right.
I’d just use Gemini 2.0 Flash if I were you - same pricing, faster, and better…
Thanks, I will check with gemini and see
