When we call Gemini 2.5 Flash Image and Gemini 2.5 Flash Image Preview, we find that the response time of the preview version is 4-5 seconds faster than that of the official version. We are not sure what causes this issue—logically speaking, shouldn’t the official version perform better than the preview version?
Hi @user3210
Welcome to the AI Forum!!
Thanks for reaching out and for sharing your observations.
Could you please share a few additional details, such as the timestamps, region, and API endpoint?
We would like to reproduce the same, Because everything appears to be working fine on our end.