Features

Cloud & Local Transcription

Typeoff supports two ASR (Automatic Speech Recognition) modes:

Cloud Mode (Default)

Audio is streamed via WebSocket to our servers for real-time transcription. This mode offers the best accuracy and speed.

• Real-time streaming transcription
• Automatic punctuation and formatting
• Falls back to local mode when offline

Local Mode

Uses a local Whisper model for fully offline transcription. No audio data leaves your device.

• Complete privacy — no network required
• Works without an internet connection
• Configurable in settings

Smart Text Insertion

Typeoff automatically detects whether your cursor is in an editable field:

• Editable field detected — text is inserted directly at the cursor position
• No editable field — a notification appears with the transcribed text, ready to copy

AI Enhancement

Typeoff can optionally process your transcription with AI to:

• Fix grammar and punctuation
• Reformat text for clarity
• Translate to another language

Translation Shortcut

Press a dedicated shortcut (Cmd+J on macOS, Alt+J on Windows) to record and automatically translate your speech to a target language. This lets you translate on demand without changing your default settings.

Recording Indicator

A compact floating capsule appears at the bottom of your screen:

• Recording — microphone icon with animated audio waveform bars and a timer. When real-time transcription is available, the live text replaces the waveform.
• Processing — a spinning loader with a progress bar while your speech is being transcribed.

The indicator never steals focus and works across multiple displays and fullscreen apps. You can enable cancel and finish buttons in settings.

Sound Feedback

Optional audio cues when recording starts and stops, configurable in settings.