There are other options like Superwhisper, local-only options like MacWhisper, and entirely open source alternatives on Github and Gradio. Whisper models + an LLM to edit raw utterances are 10x more accurate and useful than Apple dictation is these days, for me at least.