Gemini Ultra API response times?

Hey it sounds like my question is in the wrong forum however please let me explain.

I have been using GPT4 Turbo at the highest tier (I think) and get fairly decent response times from 2 - 10 seconds though at some times it’s much slower also depending on how many tokens of course.

I wonder if anyone has done a direct comparison of Gemini ultra in regards to response times covering a variety of prompt scenario’s and output lengths?

I may get the opportunity to do so but it’s my understanding that the times I am getting are pretty good given current state of tech.

I’m getting pressured for sub millisecond responses.

Also trying to avoid a potential sales job :laughing:

Really appreciate any insight and help thanks!

There’s no info on the size of GPT-4 or Gemini Ultra AFAIK so the comparison of response time might not be as valuable compared to their expertise and quality of responses.