Words arrive before the thought cools.
Streaming capture starts on key-down, cleanup happens at release, and the result lands at your cursor in a flash.
Neural dictation for people who move faster than keyboards
Stop typing.Speak in light speed.
Superlumy turns messy, natural speech into clean text anywhere your cursor blinks. Hold one key, talk, release, and watch polished words materialize across your Mac.
CursorSlackLinearNotionGmailTerminal
Superlumy CommandLive
Cleaned transcript
Let's turn the launch notes into a crisp customer update and paste it into Linear.
Capture98%
CleanupAI
Paste0.41s
System-wide
Local mode ready
0.41srelease-to-text
3.4xfaster than typing
0screenshots captured
Interface
A voice layer that feels native to tomorrow.
Raw speech becomes sharp writing.
The cleanup pass removes filler, repairs punctuation, and keeps your voice without shipping the ramble.
Local-first when the work is sensitive.
Use the on-device engine for private dictation, with no screenshots or telemetry needed to make it work.
Workflow
One gesture. Zero ceremony.
No recording windows, no context switching, no cleanup session after the thought is gone.
- 01Hold your global hotkey anywhere on your Mac.
- 02Speak naturally while Superlumy streams every phrase.
- 03Release to clean, rewrite, and paste text instantly.
$ superlumy listen --global
voice stream locked
cleanup model warmed
paste target detected
Privacy architecture
Fast does not have to mean watched.
Dictate with a local engine when you want the room sealed. Superlumy is built around the cursor, the hotkey, and your words, not invasive screen capture or background surveillance.
No telemetry required
No screenshots
Offline-ready models
Local history
Ready when you are
Give your keyboard the night off.
Install Superlumy, press the hotkey, and make every text field feel like a command line for your voice.