Thanks for posting your question in the Microsoft Q&A forum.
You can have different service tiers and hardware generations for your primary and secondary instances. It's also possible to have different vCore counts for your primary and secondary instances.
While you can have different compute resources, the secondary instance must have at least as much storage as the primary instance. This ensures that the secondary can accommodate all the data from the primary.
It's generally recommended to scale the secondary instance first before scaling the primary. ensure that the secondary instance has sufficient resources to handle the replication workload from the primary, even if you scale it down
Please don't forget to close up the thread here by upvoting and accept it as an answer if it is helpful