Built for macOS, designed for streamers who want AI without the complexity.
Local AI text-to-speech powered by the Kokoro model. 30+ voices, zero latency, zero API costs. Reads anything you throw at it.
Connect your channel and have subs, bits, and gift alerts read aloud automatically. Full chat overlay for OBS capture.
A frameless, always-on-top sine wave visualizer that reacts to your TTS audio in real time. Perfect for OBS capture.
Right-click any selected text in Chrome and send it straight to AV Stream Lab to be read aloud. No copy-pasting required.
Generate and save named voice clips as .wav files for use in OBS scenes, intro videos, or anywhere else you need them.
Everything runs on your Mac. No subscriptions, no data leaving your machine. The model runs on Apple Silicon via MLX.
Double-click the .app bundle. On first launch it installs the Kokoro model automatically — no Python setup required.
Enter your Twitch Client ID in Settings, then click Get Token. The browser handles auth and the token fills in automatically.
Capture the visualizer and chat overlay windows in OBS. Position them anywhere on your layout.
Type or paste text into the Reader, or let Twitch alerts trigger it automatically. Your voice is ready.