Wispr Flow: A new leap in voice dictation for iOS, Windows, and MacOS

Whoa. A friend of mine just told me about Wispr Flow, and it's just amazing. Here is a coupon to get a free month so you can decide for yourself if it's right for you.

It has been so annoying to voice dictate messages like “KCOR is a new algorithm” only to find it will transcribe as “cake or is a new algorithm” or something equally bad.

This it is the first system I’ve used that allows you to define your own dictionary, so you can define words like KCOR and it will get it right the first time!

In fact, I’m using Wispr Flow to dictate this entire Substack article.

It works on your iPhone with any application, and it also works on PCs and Macs. It’s pretty awesome. I’m really impressed with the accuracy.

Click the image above to get 1 MONTH of unlimited voice dictation for free.

BUGS ON WINDOWS

After you release the 2 keys, it just stays in dictation mode. This is VERY frustrating.

Release the keys, and then nothing happens. But if you open the app, you can see that the app heard you and transcribed it right, but it didn’t do the text insertion.

And then if you try to press the dictation keys again, nothing happens.

The transcript is there in the app and you can click on it and then hit your paste command (Control-v in Windows), to paste the text.

So, it’s definitely buggy at this point but it’s intermittent.

But I would imagine that they’ll fix these pretty soon.

Hitting the escape key will cancel your transcript if this happens.

With the coupon above, you get free use for a month, and after that, you get to decide for yourself whether you want to continue to use it. The cost is around $12 or so a month if you buy a yearly subscription.

It has a personal dictionary where you can add words and also correct common misspellings.

I have no affiliation with the company. I was just really impressed with the product.

Debugging Tips

• Make sure the cursor is actually BLINKING in the application where you want the text insertion before you activate Flow.

• If nothing happens, try Alt+Shift+z to paste the latest transcript.

• On the app home page, you’ll see everything that you’ve dictated. You can then select any of those lines, and you’ll see options to copy them.

  

• If you click anywhere in that box (, it will copy the transcript. You can then paste the transcript with Ctrl+V.

• I’ve had it not transcribe text after I’ve spoken it, and nothing happens. After I dictate the next segment, it’s appended to the previous segment so it’s basically in a mode where it’s appending things before it transcribes.

• While dictating, watch the Flow UI — ensure you see it listening, and only release the hotkey when done.

• After letting go of the hotkey, check: did the Flow UI show the text (confirming transcription succeeded)? If yes, did it then paste? If not, try manually Ctrl+V to see if it’s in clipboard — maybe it transcribed but failed to paste.

• Try disabling browser extensions temporarily (especially ones that affect clipboard or keyboard input) in Chrome.

• Check if NumLock (or other keyboard states) affects behavior — try toggling it off, then test again (one Reddit user found this fixed some VS Code issues) Reddit

• Update Wispr Flow to the latest version — there may be bug-fixes for certain paste behaviors.

• In Chrome, perhaps use “Paste as plain text” or make sure the editor allows programmatic pastes. In Substack’s editor, ensure it’s not in a “rich mode” that blocks injected text.

• If it still fails, switch to another text field (e.g., Notion, Word) and test whether Flow pastes there reliably. If yes, the issue is likely with Substack/Chrome integration rather than Flow per se.

Summary

It’s been a bit frustrating with these bugs.

Is it just me?

What is your experience? Let me know in the comments.