Sway - Turn spoken thoughts into clear structure.
by•
Sway is for people who think better by talking than typing.
Just speak your thoughts naturally. While walking, thinking, or reflecting.
Sway listens and turns what you say into clear summaries, key points, and actions.
No prompts. No formatting.
Just speak -> clarity.
Replies
Simplified interface looks cool, but what's the main difference with sending a voice note to ChatGPT with request to get it structured?
Sway
@vasilbo Great question actually. You can send a voice note to ChatGPT and for many cases that works.
The difference with Sway is that it’s built specifically for thinking out loud. No prompts, no chat back and forth, no instruction layer. You just speak.
It automatically structures into summary, key points, and actions consistently without you having to ask for a format each time. The goal isn’t “AI assistant”, it’s “someone who listened really well.”
And with GPT, you’re interacting with a model. With Sway, you’re capturing your thinking habit.
Tested the product. Sorry to say, its terrible.
Dint pick my voice at all nor did it get my spoken words right. Rather it assumed some random statements and showed it. This way too far away from the likes of wisprflow TBH.
Infact, the entire product could be a simple feature in any AI apps like GPT. They could possibly build this over a coffee (maybe its not too imp in their journey, but GPT gets voice input almost 80% right)
Was hoping for something that beats wispr, but NOPE.
Sway
@mithun3 Hey, thanks for taking the time to test it and for being this direct. I genuinely appreciate the honesty. That definitely doesn’t sound like the experience we’re aiming for, especially around voice pickup and accuracy. Sway is still early, and there are edge cases (mic setup, environment, language, accent, device) where things can break more than they should. If you’re open to it, I’d really like to understand what device / browser / environment you were using. This kind of feedback helps us a lot to track down issues we might be missing.
Totally fair to compare it to tools like Wispr or GPT, different approaches, different tradeoffs. We’re focused less on raw transcription and more on capturing MEANING, but none of that matters if the basics fail.
Appreciate you calling it out.