Built for Apple Silicon

Tesseract
Agent.

AI that lives on your Mac. Dictation, speech, image generation, and a local AI agent — powered by open models, processed entirely on-device. No cloud. No accounts. Just you.

See what it does

Everything runs on your Mac

Open-source AI models, optimized for Apple Silicon. No servers, no API keys, no subscriptions.

Voice Dictation

Hold a hotkey, speak, release. Your voice is transcribed and typed directly into whatever app you're using — a text editor, a browser, a chat window.

Push-to-talk On-device transcription

Text-to-Speech

Hear any text read aloud with natural-sounding voice synthesis. Consistent quality across long-form content, generated in real time on your Mac.

Natural voice No internet required

AI Agent

A local AI agent that can read, write, and edit files on your Mac. Interact by voice or text. Extensible through packages and plugins — all running privately on-device.

Tool-calling Sandboxed

Image Generation

Create images from text descriptions using on-device diffusion models. No waiting for a server, no usage limits, no content filtering by a third party.

Text-to-image No content filters

Your data never leaves your Mac

Every model runs locally on Apple Silicon. There are no servers to trust because there are no servers at all.

All AI inference runs on Apple Silicon
Works fully offline after initial setup
No accounts, no sign-ups, no telemetry
No cloud, no servers, no API calls
Open models you can inspect and replace
Sandboxed — your files stay under your control

How it works

1

Download

Get Tesseract Agent from the Mac App Store.

2

First Launch

AI models download once on first launch. This is the only time the app needs the internet.

3

Grant Permissions

Allow microphone access for dictation and accessibility access for global hotkeys.

4

Ready

Everything runs on-device from that point forward. No internet needed.

Requirements

macOS 26 or later
Apple Silicon
M1 or later