Voice-to-text on Mac has come a long way. A few years ago, your options were limited to Apple's built-in dictation or expensive professional software. In 2026, there are several genuinely capable free options — each with different strengths, limitations, and ideal use cases.
This guide covers every significant free voice-to-text option available on macOS today. We will look at what each tool does well, where it falls short, and which use cases it serves best. Full disclosure: we make Steno, which has a free tier, so we are naturally biased. But we will try to give every option a fair assessment, because the best tool is the one that fits your specific needs.
Apple Built-in Dictation
Every Mac ships with dictation built into the operating system. You can activate it through System Settings under Keyboard, and trigger it with a double-tap of the Function key (or whatever shortcut you configure).
Strengths
- Zero installation required — it is already on your Mac
- Deep system integration — works in most native macOS applications
- On-device processing option in macOS Ventura and later, meaning audio stays on your Mac
- Supports voice commands for editing ("delete that," "new paragraph")
- Completely free with no usage limits
Limitations
- Accuracy lags behind modern AI transcription models like Whisper, particularly for technical vocabulary, accented English, and noisy environments
- The activation flow requires switching focus — a floating dictation indicator appears and can interfere with some applications
- On-device processing is noticeably less accurate than the cloud-based option
- No hold-to-speak model — you toggle dictation on and off, which requires explicitly ending the session
- Limited language model means more frequent errors with homophones and unusual sentence structures
Best For
Users who want basic dictation without installing anything and do not mind occasional accuracy issues. Good for short messages and simple text input where perfect accuracy is not critical.
Google Docs Voice Typing
Google Docs includes a built-in voice typing feature accessible through Tools, then Voice typing (or Command-Shift-S). It works in Chrome on macOS.
Strengths
- Good accuracy, powered by Google's speech recognition
- Supports over 100 languages and dialects
- Voice commands for formatting ("bold," "italics," "new line")
- Free with a Google account
Limitations
- Only works inside Google Docs in Chrome — you cannot use it in other applications
- Requires an internet connection at all times
- Audio is processed by Google's servers with their standard data policies
- No way to use it for composing emails, Slack messages, or text in other apps without copy-pasting
Best For
Users who do most of their writing in Google Docs and want a built-in option without any additional software.
Web-Based Speech-to-Text Tools
Several web applications offer free speech-to-text using the Web Speech API built into Chrome. Tools like Speechnotes, Dictation.io, and TalkTyper provide browser-based dictation interfaces.
Strengths
- No installation required — just open a browser tab
- Generally decent accuracy through Chrome's speech recognition
- Some offer real-time transcription with minimal delay
- Free, often with no account required
Limitations
- Tied to a browser tab — you dictate into the web app and then copy text elsewhere
- Accuracy depends on Chrome's speech recognition, which varies by language and accent
- Cannot inject text directly into other applications
- Require a constant internet connection
- Privacy varies — some tools send audio through their own servers in addition to Google's
Best For
Quick, occasional transcription tasks where you do not mind copying and pasting the result.
Whisper.cpp and MacWhisper
OpenAI's Whisper model has been open-sourced, and several projects have made it available for local use on Mac. Whisper.cpp is a C++ implementation that runs the Whisper model locally, and MacWhisper provides a native Mac interface for it.
Strengths
- Completely offline — all processing happens on your Mac
- Excellent accuracy, especially with the larger model sizes
- No data leaves your machine, providing the strongest possible privacy
- Whisper.cpp is free and open source
Limitations
- Requires downloading large model files (the best-quality model is several gigabytes)
- Processing speed depends on your hardware — the base model is near real-time on Apple Silicon, but larger models take considerably longer
- No system-wide text injection — you transcribe audio files or record into the app, then copy the result
- Not designed for quick, inline dictation — better suited for transcribing recordings
- MacWhisper is free for basic features but charges for Pro features
Best For
Users who need to transcribe recordings (meetings, interviews, lectures) and prioritize complete privacy over real-time inline dictation.
Steno Free Tier
Steno offers a free tier that provides the core dictation experience: hold the hotkey, speak, release, text appears at cursor. The free tier has a monthly usage limit but includes the same AI-powered transcription accuracy as the paid plan.
Strengths
- Works system-wide — dictate into any application that accepts text
- Hold-to-speak model is the fastest activation method available
- Powered by Groq's Whisper API — same accuracy as the best AI transcription
- Native macOS app at 1.7MB, runs in the menu bar with negligible resource usage
- Automatic punctuation and formatting
- No account required to start using
Limitations
- Free tier has a monthly usage cap — heavy users will hit the limit
- Requires internet connection for transcription
- Advanced features (smart rewrite, meeting mode) are Pro-only at $4.99/month
Best For
Users who want the best possible accuracy for inline dictation across all applications, and whose usage fits within the free tier limits. Also ideal for users who want to try high-quality AI dictation before committing to a paid plan.
Comparison Summary
The right free voice-to-text tool depends on your primary use case.
If you want something that just works without installing anything and you primarily use native Mac apps, Apple's built-in dictation is the path of least resistance. The accuracy is acceptable for casual use, and it is genuinely zero-effort to set up.
If you live in Google Docs, the built-in voice typing is excellent and requires no additional software. The accuracy benefits from Google's speech recognition, and the voice commands for formatting are genuinely useful for long documents.
If privacy is your absolute top priority and you are willing to sacrifice real-time dictation for complete data isolation, Whisper.cpp running locally is the most private option possible. Nothing leaves your machine.
If you want the best combination of accuracy, speed, and system-wide usability in a free tool, Steno's free tier provides AI-powered transcription that works everywhere on your Mac with the fastest possible activation model. The monthly cap is generous enough for light to moderate dictation use.
The Case for Upgrading
Free tools cover the basics, but if voice-to-text becomes a core part of your workflow, the limitations of free options will eventually become friction. Usage caps, reduced accuracy, app-specific restrictions, and copy-paste workflows all add up over time.
Steno Pro at $4.99 per month removes the usage cap and adds features like smart rewrite (automatic tone adjustment) and meeting mode (continuous transcription). For users who dictate daily, the time saved easily justifies the cost — most users report saving 30 to 60 minutes per day compared to typing everything manually.
But start with the free tools. Try Apple's built-in dictation. Try Steno's free tier. Experience the difference between basic and AI-powered transcription firsthand, and then decide if upgrading makes sense for your workflow.
You can download Steno's free tier in seconds — the entire app is a 1.7MB download.