Sumi – Local AI that cleans and adapts your speech to text

Summary: Sumi converts spoken words into polished text by removing filler words and adjusting tone based on the app context. It supports over 100 languages, runs locally on macOS and Windows, and offers meeting recording and voice editing features without cloud processing.

What it does

Sumi transcribes speech while cleaning grammar and filler words, tailoring output for apps like Gmail or Slack. It records calls in the background and allows voice commands to edit selected text.

Who it's for

Developers and users needing accurate, context-aware transcription without cloud reliance, especially those handling accented speech or multiple languages.

Why it matters

It eliminates the need for cloud-based transcription services, reducing latency and privacy risks while improving transcription quality locally.