Which statement correctly compares SQS standard and FIFO queues?

Standard queues offer at-least-once delivery with best-effort ordering. They scale to very high throughput, but they can deliver a message twice or out of order. So "Standard queues guarantee strict ordering; FIFO queues deliver in any order" is wrong.

FIFO queues keep messages in order within each message group ID. They drop duplicates sent within the deduplication window. Their throughput is lower than standard queues, so "FIFO queues have no throughput limits, unlike standard queues" is wrong.

Consumers should still be idempotent: processing the same message twice must give the same result. Why? A consumer can crash after processing a message but before deleting it, and then SQS delivers it again. So "Both deliver each message exactly once, so consumers need no idempotency" is wrong.