Breeze Mac Whisper
Professional Speech Recognition · macOS
Breeze Mac Whisper is a deeply optimized macOS speech recognition tool built on OpenAI Whisper. It's faster than native Whisper, solves the inconsistent Simplified/Traditional output problem common in other apps, and uniquely supports speaker identification and live microphone transcription.
Version: 3.58.0a
Real Product Screens
Every screenshot is an actual run of the app — nothing staged.
Flexible Model Selection
From Tiny to Large v3 Turbo — trade off speed and accuracy, load models on demand.
Live Subtitle Overlay
Microphone transcription plus a draggable, resizable floating window that never interrupts recording.
Bilingual Subtitle Editor
Timeline, segmentation, and speaker labels — fine-tune line by line, export to SRT / TXT in one click.
Local Model Management
Accelerated mirror downloads, every model runs fully offline — nothing is uploaded.
Detailed Settings
Segmentation mode, export format, light/dark theme — all configurable to your workflow.
Features
Faster Than Native Whisper
Internally optimized for speed — transcription finishes significantly faster than running OpenAI Whisper natively.
Fixed Simplified / Traditional Output
Most Whisper apps produce inconsistent zh/tw output. Breeze lets you lock output to Simplified or Traditional — every time.
Speaker Identification
Identifies and labels multiple speakers in a recording. Native Whisper has no speaker diarization — Breeze does.
Live Microphone Transcription
Real-time transcription from your microphone — ideal for meetings, lectures, and live captioning.
Bilingual Subtitles & Translation
Transcription and translation displayed side by side. Set a target language and generate dual-language subtitles in one step.
Floating Subtitle Window
A subtitle overlay that floats above any app — perfect for watching foreign-language video in real time.
Fully Local — Private by Design
Everything runs on your Mac. No audio or video ever leaves your device.
Apple Neural Engine Acceleration
Fully utilizes Apple Silicon's Neural Engine for inference — your Mac's hardware working at full capacity.
Who It's For
What Users Say
Genuine feedback from paying customers
I have to give this software huge praise — it's probably the only Mac Whisper app truly built for Chinese users. Dense interview audio used to get cut mid-sentence; now it segments by meaning automatically. Such a relief.
I process tons of interview footage for content work. Other tools used to mix up two speakers' lines. This one runs buttery smooth on Mac, and best of all it can parse a video link directly — drop in the URL and get an srt file minutes later.
Bought it mainly for course recordings and interview material. The Chinese UI is convenient, recognition is faster than expected, and it handles long audio fine. Speaker identification is genuinely accurate.
Great for video subtitles — tried it on a few English videos, recognition held up well and timestamps lined up. Support also replied quickly.
The interface is simple, not cluttered, and model downloads are fast. Importing files, running recognition, exporting subtitles — all intuitive. Overall a solid experience.
Questions or collaboration?
Reach out via email or the contact form. We typically respond within 1–2 business days.
Contact Us