Speech to Text

Convert your voice to text in real-time using on-device AI. No internet required, no data sent anywhere.

How It Works

Dial8 uses Whisper, an open-source speech recognition model, running natively on your Mac's Apple Silicon. When you speak, your voice is processed locally and the transcribed text is inserted wherever your cursor is.

Recognition Models

Choose from two Whisper models based on your needs:

Models are downloaded on demand during setup and stored locally on your Mac.

Language Support

Dial8 supports automatic language detection or you can lock to a specific language for improved accuracy. Supported languages include English, Chinese, German, Spanish, Russian, Korean, French, Japanese, Portuguese, Turkish, Polish, Italian, Vietnamese, Dutch, Persian, Arabic, and more.

Input Modes

Push-to-Talk

Hold the Option key (or your configured hotkey) while speaking. When you release, the audio is transcribed and typed into your active app.

You can press Space while holding the hotkey to "lock" the recording — it continues without needing to hold the key. Press the hotkey again to stop.

Toggle Mode

Press the hotkey once to start recording, press again to stop and transcribe.

Streaming Mode

For real-time transcription, streaming mode detects natural pauses in your speech and transcribes segments incrementally. Text appears as you speak rather than waiting until you stop.

AI Text Cleaning

Optionally, Dial8 can use AI to clean up your transcription — removing filler words, fixing punctuation, and adapting tone to the app you're typing in. When enabled, the AI detects whether you're in a professional app (Slack, email) or casual app (iMessage, Discord) and adjusts formatting accordingly.

Enable this in Settings > AI Rewrite.

Text Replacements

Set up automatic find-and-replace rules for common corrections — like fixing name spellings or expanding abbreviations. For example, "PM" can automatically become "product manager."

Configure these in Settings > Text Replacements.