While analyzing up to 50 images at same time the latency of model increases exponentially up to 60 seconds why and how to solve this?
It seems obvious if you increase the number of images, it’s going to take few more seconds. Please try to switch to low latency models like 1.5-flash or 1.5-flash-8B and see if you are able to save some time.
Thank you @Govind_Keshari , Sure i will try this approach but can you explain me this scenario when i am requesting gemini asynchronously over 100 images parallelly , I have put time logs to check the time of response for every image is increasing exponentially can you explain why this ?
It seems like you are hitting some kind of throttling.
Does the same thing happen if you do it with 20 or 40 images?