Hey 🙂 I’m curious about the performance calculations for the search and where they come from. I’m interested because when I perform a search, the results are listed as taking 20ms but the network request takes 313ms. Is this figure the number purely for the Typesense search function to finish executing, and nothing else? I want to showcase the speed of the search so I’d like to have some data to prove this number on my end. Is that possible?

Yes the 20 ms is raw computation time. End to end latency will depend on network latency. If the Typesense server is another continent then that latency will be much larger. Typesense Cloud allows you to create a geographically distributed cluster which helps with network latency.

Thanks Kishore! Good to know. In my case I have to compare with another system so in order to compare like with like I will end up taking the value of the network request resolution, but the improvement is still astonishing!