If you’re looking for a way to turn speech into text on your Mac, you’ve got more options than ever. But they vary significantly in terms of privacy, cost, accuracy, and workflow integration. Here’s a practical look at the main contenders in 2026.
Apple Dictation (built-in)
Price: Free Privacy: On-device (Enhanced Dictation) or cloud Best for: Quick, casual dictation
Every Mac ships with dictation built into the operating system. You can enable it in System Settings > Keyboard and trigger it with a keyboard shortcut (double-tap the Fn key by default).
Apple’s on-device dictation has improved significantly over the years, especially on Apple Silicon. It handles conversational speech reasonably well and supports voice commands for punctuation (“period,” “new paragraph”).
Limitations: It works well for short bursts of text, but longer dictation sessions can be inconsistent. There’s limited ability to customize or extend it, and the transcription quality for technical terminology or unusual proper nouns can be hit or miss. It also doesn’t do any post-processing to clean up filler words or improve sentence structure.
OpenAI Whisper
Price: Free (open source) / API pricing for cloud version Privacy: Local (self-hosted) or cloud (API) Best for: Developers comfortable with command-line tools
Whisper is OpenAI’s open-source speech recognition model. You can run it locally using the open-source version, which gives you excellent transcription quality with full privacy. The cloud API is also available if you prefer not to manage the infrastructure yourself.
Limitations: Whisper is a transcription tool, not a dictation tool. It works on audio files rather than live speech, so it doesn’t integrate into a real-time typing workflow without additional tooling. Running it locally requires Python and some technical setup. The larger models need significant GPU memory for reasonable performance.
Dragon by Nuance
Price: Starting at ~$200/year Privacy: Cloud-based Best for: Professional transcription, especially medical and legal
Dragon has been the gold standard for professional dictation software for decades. It offers deep vocabulary customization, voice commands for controlling your computer, and specialized editions for industries like healthcare and law.
Limitations: Dragon is expensive and has shifted to a subscription model. The macOS version has historically lagged behind the Windows version in features and stability. It’s cloud-based, which means your audio is sent to Nuance’s (now Microsoft’s) servers for processing. For the casual user, it’s overkill.
Otter.ai
Price: Free tier / $16.99/month Pro Privacy: Cloud-based Best for: Meeting transcription and collaboration
Otter.ai is primarily a meeting transcription tool. It joins your Zoom, Google Meet, or Teams calls and produces timestamped, speaker-labeled transcripts. It’s excellent at what it does: collaborative note-taking with searchable transcripts.
Limitations: Otter is designed for meetings, not general-purpose dictation. It’s entirely cloud-based, so all your audio goes to their servers. The free tier is limited, and the pricing adds up if you need it for a team. It’s not really meant for typing text into an email or document.
Voxr
Price: Free Privacy: Fully local (on-device) Best for: Privacy-conscious users who want fast, everyday dictation
Voxr takes a different approach. It’s a native macOS app that combines Apple’s speech recognition with on-device AI to produce clean, polished text from your voice. Everything runs on your Mac, with no cloud, no subscriptions, and no data leaving your machine.
You speak, Voxr transcribes and processes your words through a three-stage pipeline, and the result is pasted directly into whatever app you’re using. The LLM step handles punctuation, filler word removal, and light cleanup that raw speech recognition typically misses.
Limitations: Processing speed depends on your hardware (Apple Silicon recommended). The AI processing adds a few seconds of latency compared to raw dictation.
How they compare
| Feature | Apple Dictation | Whisper | Dragon | Otter.ai | Voxr |
|---|---|---|---|---|---|
| Price | Free | Free/API | ~$200/yr | $17/mo | Free |
| Privacy | Mostly local | Local or cloud | Cloud | Cloud | Fully local |
| Real-time | Yes | No | Yes | Yes | Yes |
| Post-processing | No | No | Limited | Yes | Yes (LLM) |
| Offline | Yes | Yes (local) | No | No | Yes |
| Setup difficulty | None | Moderate | Low | Low | Low-moderate |
Which one should you use?
If you just need quick dictation for short messages, Apple’s built-in Dictation is already on your Mac and works fine.
If you’re transcribing recorded audio or need the highest raw accuracy, Whisper (self-hosted) is hard to beat.
If you’re in a regulated industry and need specialized vocabularies, Dragon is still the professional choice despite the price.
If your main need is meeting transcription with team collaboration, Otter.ai is purpose-built for that.
If you want everyday voice-to-text that’s private, free, works offline, and produces clean output, Voxr is worth a look. Learn more about why local AI matters for voice-to-text.