Mimic Wispr Flow but 100% local

Posted by Present_Ride6012@reddit | LocalLLaMA | View on Reddit | 7 comments

Created an menu bar App using Whisper models under the hood running 100% locally. Not sure how to sign my App properly yet.

\~80MB memory usage for Large v3 turbo model on a M4 MBA. Enjoy transcribing and interacting with LLMs with NO token limits.

https://github.com/t2o2/local-whisper