Be My Ears provides system-wide live captions for any audio playing on a Mac. The app captures sound from the Mac’s speaker and converts it into real-time subtitles, working with Zoom, Teams, Google Meet, Skype, web browsers, and essentially any application that produces audio. This universal compatibility distinguishes it from meeting-specific transcription tools.
A key feature is complete on-device processing using Apple’s CoreML and Speech framework. Audio never gets uploaded to external servers, making it suitable for confidential business conversations. The transcription runs entirely on the Mac using Apple’s machine learning models.
Multilingual support covers over 20 languages including Chinese (Simplified and Traditional), English (American, British, and Indian variants), French, German, Italian, Japanese, Korean, Portuguese, Spanish, Russian, and Arabic. The app automatically detects the language being spoken. For macOS 15 Sequoia users, real-time on-device translation translates live audio into a chosen language with sub-second delay, all without internet connectivity.
Beyond live captions, the app records and exports both transcriptions and audio (M4A format). An “Ears Care” function monitors audio levels and provides reminders when sound is too loud, reflecting the developer’s accessibility focus.
The free version provides unlimited core transcription with no time limits—no freemium limitations or artificial restrictions.
System requirements: macOS 13.0 Ventura or newer. Optimized for Apple Silicon’s Neural Engine (M1, M2, M3, M4 series). Intel-based Macs are supported on compatible macOS versions. Initial setup requires downloading Apple’s language models (up to 30 minutes depending on connection).
Limitations: Each language model requires a separate download. Users working with multiple languages need multiple model downloads. The app depends on Apple’s built-in speech recognition, so transcription accuracy varies by language and audio quality.
Alternatives: macOS built-in Live Captions (macOS Ventura+) provides basic system-wide captioning. Whisper Transcription offers alternative on-device transcription using OpenAI’s Whisper model. Meeting-specific tools like Otter.ai or Krisp provide transcription with cloud processing.
Suitable for users who take regular video calls, work with international teams, are learning new languages, or need accessibility support—particularly those who prioritize privacy and prefer completely on-device processing.
