Best way to transcribe audio to text? (Podcast episodes)

Preferably FOSS or similar but open to anything that's free.