Real-time audio to text for any Mac app

Powered by local AI. No internet required. Your data never leaves your Mac.

Requires Apple Silicon Mac (M1+) · macOS 14.2+ · 16GB RAM

Works with any app

Chrome
Safari
Spotify
Zoom
Discord
Slack
YouTube
Podcasts

Features

Everything you need, nothing you don't

Built from the ground up for macOS. Lightweight, private, and incredibly capable.

Real-time Transcription

Watch words appear as they are spoken. Powered by state-of-the-art Qwen3-ASR model running entirely on your Mac.

100% Private

All processing happens locally. Your audio never leaves your device. Zero data collection.

52 Languages

From English to Cantonese, Japanese to Arabic. 22 Chinese dialects included.

Source Filtering

Choose which audio source to transcribe. Works per-app with precise control.

Light & Dark Mode

Follows your system appearance seamlessly. Looks great any time of day.

Adjustable Opacity

Fine-tune subtitle window transparency to blend with your workflow.

How It Works

Three steps. That's it.

No sign-up, no API keys, no cloud. Just open and go.

Open any audio source

Play a video, join a meeting, stream a podcast, or use any app that outputs audio.

Click to start capturing

Select the audio source from the menu bar and hit start. AudioTextLayer begins listening instantly.

Watch subtitles appear

A floating window shows real-time transcription. Position it anywhere on your screen.

Product

Designed to stay out of your way

A clean menu bar popover and a floating subtitle window. Nothing more.

Menu bar popover with audio source selection and real-time transcript view.

Comparison

AudioTextLayer vs Cloud Services

See why local AI transcription is the smarter choice.

Feature AudioTextLayer Otter.ai YouTube Captions Zoom Transcription
Privacy 100% Local Cloud Cloud Cloud
Languages 52 1 Varies 12
Works with any app Yes No No No
Real-time Yes Yes Varies Yes
Cost Free $16.99/mo Free Paid plan
Internet required No Yes Yes Yes

Languages

52 Languages. One Click.

From English to Cantonese, from Japanese to Arabic -- we've got you covered.

English 普通话 (Mandarin) 粤語 (Cantonese) 日本語 (Japanese) 한국어 (Korean) Français Deutsch Español Português Italiano Nederlands Русский العربية हिन्दी ไทย Tiếng Việt Bahasa Indonesia Bahasa Melayu Türkçe Polski Čeština Română Magyar Ελληνικά Svenska Dansk Norsk Suomi עברית বাংলা فارسی ഈാളം

FAQ

Frequently Asked Questions

AudioTextLayer is a macOS menu bar application that captures audio from any app and converts it to text in real-time using a local AI model. It displays live subtitles in a floating window you can place anywhere on screen.
AudioTextLayer captures system audio using macOS audio routing, then processes it through a locally-running Qwen3-ASR AI model. Everything happens on your Mac -- no internet connection or external servers are involved.
Yes, 100%. Audio is processed entirely on your Mac. No data is ever sent to any server. No analytics, no telemetry, no cloud. Your conversations stay yours.
You need an Apple Silicon Mac (M1 or later) with at least 16GB of RAM, running macOS 14.2 (Sonoma) or later. The AI model leverages the Neural Engine for optimal performance.
Yes. AudioTextLayer is currently in free beta. Download and use all features at no cost.
AudioTextLayer supports 52 languages including English, Mandarin, Japanese, Korean, French, German, Spanish, Portuguese, Arabic, Hindi, and many more. It also includes 22 Chinese regional dialects.
No. After the initial download, AudioTextLayer works fully offline. The AI model runs locally on your Mac's Neural Engine. No internet connection is needed.
AudioTextLayer uses the state-of-the-art Qwen3-ASR model, which delivers excellent accuracy across all supported languages. Performance is comparable to leading cloud services, with the added benefit of complete privacy.

Start transcribing in seconds

Download AudioTextLayer and experience real-time transcription on your Mac.

Download for Mac — Free Beta

Requires Apple Silicon Mac (M1+) · macOS 14.2+ · 16GB RAM