Private, local text-to-speech

Your Mac, out loud.

SayIt turns selected text into natural speech with open models running entirely on Apple silicon. No cloud, no accounts, no analytics — your words never leave the machine.

Download for macOS View on GitHub

macOS 15+ · Apple silicon · Free & open source

A first look

Press a hotkey. It reads.

The player floats in your menu bar — follow the spoken word, scrub the waveform, change the speed. Models download once from Hugging Face and speak offline from then on.

SayIt's menu-bar player reading a paragraph aloud, with a waveform, playback controls, and the current word highlighted — 01 **Lives in your menu bar** Follow each word, scrub the waveform, change the speed.

SayIt's model library listing open MLX speech models with their size, speed, languages, and license — 02 **Open models, one click away** Kokoro, Qwen3, OmniVoice and more — download once, run offline.

Free, open source, and yours.

Download for macOS