Medical Dictation for Mac: A Physician's Guide
Dragon NaturallySpeaking dominated medical dictation for decades — on Windows. When Nuance killed Dragon for Mac in 2018, physicians who preferred Apple hardware were left with a gap: Apple Dictation lacks medical vocabulary depth, cloud-based AI scribes introduce HIPAA complexity, and most alternatives are Windows-first afterthoughts.
If you're a physician on a Mac looking for dictation that actually works for clinical documentation, here's what to look for — and where the options stand.
What physicians need from a dictation tool
Not all dictation tools are built for clinical work. General-purpose voice-to-text might handle emails fine, but medical documentation has specific demands:
- Medical vocabulary accuracy. Specialty terminology, drug names, dosages, procedures, anatomical terms. A tool that stumbles on “metformin” or “cholecystectomy” creates more work than it saves.
- Speed. Clinical thinking doesn't wait. Dictation needs to keep pace with your thought process, not slow you down with lag or buffering.
- Privacy and HIPAA. Patient names, diagnoses, medications, and history are Protected Health Information. Any tool that sends PHI through third-party servers requires a Business Associate Agreement — and introduces risk even with one.
- Works where you work. Your EHR's web interface, Apple Notes, email, referral letters. Not locked to a single platform.
- Mac-native. Not a Windows port running through Rosetta. Not a browser-only web app. Something built for macOS.
The three categories of medical dictation
Medical dictation tools fall into three broad categories. Understanding the tradeoffs helps you pick the right one.
Cloud AI scribes
Tools like Freed, DeepScribe, and Abridge listen to patient encounters and generate structured SOAP notes automatically. Powerful, but expensive ($200–500/month), cloud-dependent, and require transmitting full audio of patient conversations to remote servers. A BAA is essential.
Cloud dictation
Tools like Dragon Medical One and VoiceboxMD provide real-time speech-to-text via cloud processing. Dragon Medical One ($79–99/month) has been the industry standard for years. Accurate, with deep EHR integrations — but your audio travels to servers you don't control.
Local dictation
Speech recognition runs entirely on your device. No server, no PHI transmission, no BAA required. Resonant sits here — processing everything on your Mac using on-device AI models. Less expensive, fully offline-capable, and structurally private.
Why local processing matters for physicians
When you dictate a clinical note using a cloud-based tool, your patient's name, diagnosis, medication list, and history travel through third-party infrastructure. Most cloud dictation services handle this responsibly — but the data still leaves your device.
A Business Associate Agreement shifts some liability, but it doesn't eliminate risk. Data breaches in healthcare are expensive and common. The average cost of a healthcare data breach reached $10.93 million in 2023, according to IBM — the highest of any industry.
Local processing eliminates the vector entirely. If audio never leaves your Mac, there's nothing to breach, subpoena, or mishandle. You don't need a BAA because there's no third party involved. Your patient data stays exactly where it should — on your machine, under your control.
How Resonant works for physicians
Resonant runs state-of-the-art speech recognition models directly on Apple Silicon. The Neural Engine in M-series chips provides the compute needed to rival cloud accuracy — without a network connection.
- Dictate into any app. Your EHR's web interface, Apple Notes, email, referral letters — anywhere you can type.
- Works offline. Dictate in your office, at the hospital, on rounds, or anywhere else. No Wi-Fi needed.
- No account required. Download, install, start speaking. No sign-up, no server configuration, no IT department involvement.
- Fast. Text appears as you speak. No cloud round-trip latency.
- Medical terminology. The underlying speech models handle clinical language, drug names, and procedures without custom vocabulary setup.
What Resonant doesn't do
We'd rather be upfront about what Resonant isn't:
- Not an AI scribe. Resonant doesn't listen to patient encounters and generate notes. It's a dictation tool — you speak, text appears. You control the structure.
- No deep EHR integrations. No proprietary plugins for Epic or Cerner. Resonant works with any app via standard text input — including your EHR's web interface.
- Mac only. No Windows version, no iOS app (yet). Built specifically for macOS on Apple Silicon.
For physicians who want an AI scribe that generates SOAP notes from patient conversations, tools like Freed or DeepScribe may be the right fit. For physicians who want fast, private dictation they control — Resonant is built for that.
Read how Resonant handles your voice data →
Frequently asked questions
Is Resonant HIPAA compliant?
Resonant's architecture inherently minimizes HIPAA risk because no Protected Health Information is transmitted or stored by Resonant. All speech recognition runs on your Mac — audio never leaves your device. Since there's no third-party data processing, no Business Associate Agreement is needed. You remain the sole controller of your patient data.
Can Resonant handle medical terminology?
Resonant uses state-of-the-art speech recognition models that handle medical terminology, drug names, and procedure terms. The underlying models are trained on broad datasets that include medical language. While it doesn't have a dedicated medical vocabulary add-on, physicians find it handles clinical dictation well out of the box.
How does Resonant compare to Dragon Medical One?
Dragon Medical One is cloud-based, costs $79–99/month, and sends your audio to Nuance's servers for processing. Resonant runs entirely on your Mac with no cloud component, is significantly less expensive, and ensures your patient data never leaves your device. Dragon offers deeper EHR-specific integrations; Resonant works with any app via standard text input.
Do I need an internet connection to use Resonant?
No. Resonant runs 100% on-device. You can dictate clinical notes offline — in your office, at a hospital with unreliable Wi-Fi, or anywhere else. No network connection is needed for dictation.
What Mac do I need to run Resonant?
Resonant requires a Mac with Apple Silicon (M1 chip or later). This includes all MacBook Air, MacBook Pro, iMac, Mac mini, Mac Studio, and Mac Pro models released from late 2020 onward. The Neural Engine in Apple Silicon provides the compute needed for real-time on-device speech recognition.
Try Resonant free
Private voice dictation for Mac and Windows. 100% on-device, no account required. Download and start speaking in under a minute.