I have a

LazyColumn

with around ~75 items, including `stickyHeader`s, inside a bottom sheet, and the performance of

LazyListState.animateScrollToItem()

is noticeably bad. The animation is not smooth; it appears to just show a few frames before landing at the target index. Scrolling through the list manually otherwise works well enough. The items in the list have some state associated with them (they can be “favourited” (via a heart icon) or “toggled” (via a switch). They have other components which are visible (or interactive) conditionally. So they are perhaps relatively complex. Should I expect

animateScrollToItem

to be performant in this case? I’m wondering if others have encountered a similar situation. Thanks for the feedback.

do you notice a difference between animating programmatically and touch-flinging at a similar speed?

No, at least to me it appears similar when I animate vs fling.

Are you running a release build? Do you see this behavior the first few times and then it settles?

Yes I started testing with a release build after reading that that made a difference. I’m not sure I noticed much difference in release vs debug, but I’ll check again. What to make of the suggestion of setting

minifyEnabled true

and using

proguard-android.txt

instead of

proguard-android-optimize.txt

? https://www.reddit.com/r/androiddev/comments/n61qxh/jetpack_compose_lazycolumn_laggy/h8b9c8m/?utm_source=reddit&utm_medium=web2x&context=3 I did test these changes but did not observe a noticeable difference.

If you notice it gets smoother after a few scrolls in the same run of the app you might be seeing jit compilation. We should be publishing some new docs soon on the baseline profile installation process we use such that end users generally never see this, and you can trigger the ahead of time compilation process manually for development testing

Thanks for the feedback. I will probably circle back to working on this next week, since the component otherwise needs some input from a teammate.

just putting myself here - because I came looking for the exact same problem.