Voice input for Cursor?
byβ’
Has anyone been successfully talking to Cursor? I started using Superwhisper but having to copy-paste the generated transcription into Cursor is a chore. Looking for something that just works.
7.8K views
Has anyone been successfully talking to Cursor? I started using Superwhisper but having to copy-paste the generated transcription into Cursor is a chore. Looking for something that just works.
Replies
Coding Rooms
On mac you can use Tight https://tight.sh/. It's built for developers and automatically adds code you highlight and things you point your mouse at to your prompt. Makes it feel more like pair programming.
CamelAI
I use @MacWhisper in Cursor. You can activate it in any text field when holding down the fn key. It replaces the system dictation. Works great!
Product Hunt
@vercantezΒ you can't mention files using voice, right?
CamelAI
@stevebΒ You can't @ files but that's not much of a problem for me since I mostly use Composer agent mode which will search for the file and add it to the context automatically if you mention it like "look at views.py to see how we do this first"
Intersect Labs
@vercantezΒ Thanks Miguel! Will check this out.
Raycast
Try Talktastic.
Intersect Labs
@chrismessinaΒ Awesome, thanks! Will do.
Product Hunt
@chrismessinaΒ curious if you'd say it performs better than @MacWhisper? Been looking at trying one but I'd be using it mostly for in-apps or recording notes/meetings.
Raycast
@gabeΒ TalkTastic does two things better than MacWhisper:
It improves your dictation to handle grammar and other disfluencies.
It rewrites the dictation transcript to be more suitable for the current visible app β i.e. it takes a screenshot and uses that as context for its LLM rewrite.
I'd experiment with both and see which works better for you!
@gabeΒ @chrismessinaΒ Actually not a fan of the whole screenshot of app thing, especially after reading their privacy policy. Immediately got rid of it after I seen that feature.
Raycast
@williamiparkΒ what'd you see in the privacy policy that concerned you?
The app uses a screenshot to set the context for its transcription; that's how it adjusts the tone/content to be more appropriate.
But absolutely, if that's something you're not cool with β I wouldn't use it.
It's Awesome, Thanks for mentioning it @chrismessinaΒ
Product Hunt
Product Hunt
I started using @Wispr Flow with Replit and then migrated to using it more generally - really slick and fast implementation!
Wispr Flow
@rajiv_ayyangarΒ 100% agree
Product Hunt
@tanaykothariΒ π π
I use the native voice commands on Mac. Hitting CTRL-CTRL activates and deactivates it. The biggest pain is that for technical terms it does a poor job of getting them right. i.e. SQL != "sequel" and so on. You do have to go back and edit a bit, but it is still a huge time saver.
AND I get to feel like I am living in the future.
Intersect Labs
@ammonbrownΒ Thanks! Also for the video :)
Navi for Vision Pro
You can use MacWhisper , the free version has dictation as well (I make it)
Wispr Flow
Hey, founder of @Wispr Flow here. We are working on improvements to voice dictation in Cursor-like apps.
If there are any feature requests you have, feel free to drop them in here!
If you try out the Whisper Keyboard, you'll see that there's no voice input software that's more convenient than it is.
@river_rayΒ Definitely!
Yeah, I ran into the same pain with Superwhisper. Have you tried Wispr Flow or Dictator? They plug straight into text fields so you donβt need to copy-paste every time.
@agordhandas superwhisper is good , app on iOS and Mac , and if you enable it in Settings and then it type in any text Field option-plus space key. And I wrote this comment, the last part of it actually using superwhisper. The first part I tried to type myself.