For larger data its normally recommended to push that data to R2 and then passing an ID in the queue and pulling that info from R2 (or KV even)