I believe the consensus among mathematicians is that Fermat either didn't have a proof or had an incorrect proof.

Any statement of the modern proof would require hundreds of pages and results not available to Fermat. So it's likely Wiles didn't rediscover the proof but simply discovered it.