The Verge writer has tested several speech-to-text applications including WisprFlow, Monologue, Spokenly and Handy to assess their utility for drafting emails and Slack messages. These tools have improved significantly in processing spoken input, though they often default to overly formal formatting and incorrectly add periods to informal text messages. The author notes that even the fastest and cheapest large language models now understand and process speech with remarkable accuracy, making them effective for quick writing tasks despite minor stylistic quirks.
This development matters because it shifts the focus of artificial intelligence gadgets from complex devices like smart rings to practical software tools that improve daily communication workflows. The core value lies in the immediate availability of high-quality dictation capabilities without requiring expensive hardware upgrades. Users can now generate text faster while accepting that some applications may need manual editing to correct tone or punctuation errors.
- WisprFlow, Monologue and Spokenly are among the tested apps
- Models now process speech accurately even at low cost
- Manual correction is often needed for punctuation and tone
