Your voice,perfectly transcribed

Transform speech into polished text with AI that runs entirely on your Mac. No cloud, no subscriptions, no limits.

Trusted by developers at

Run your transcription locally

Extend your workflow with local AI. Run inference for speed and privacy.

Your Mac got powerful.

Modern Apple Silicon can run real AI inference. Dicta uses that local power for instant, private transcription.

Cloud stays optional.

Keep your existing workflows. Use cloud when you need scale, local when you need speed and privacy.

Zero latency.

Running inference on-device means instant results. The kind of speed no cloud service can match.

Privacy is built-in.

Your voice data never leaves your device. Local processing means complete control and trust.

See it in action

Watch Dicta in action

See how fast and accurate local transcription can be. No internet required, no data leaves your Mac.

Watch demo video

<100ms
Latency
50+
Languages
100%
Private
Free
Forever

Key benefits

Instant, private transcription.

Near-zero latency and full data privacy.

Custom vocabulary learning.

Train on your names, terms, and technical jargon.

Global keyboard shortcuts.

System-wide hotkeys that work in any app.

Voice-activated snippets.

Say trigger words, expand to full text blocks.

50+ languages supported.

Transcribe across languages with Whisper AI.

Advanced capabilities

Power features for power users

Dicta goes beyond basic transcription. Customize it to fit your workflow perfectly.

Custom Vocabulary

Train Dicta on your names, technical terms, and domain-specific jargon for better accuracy.

Voice Snippets

Say trigger words to instantly expand into full text blocks. Perfect for signatures and templates.

Multi-Language

Transcribe in 50+ languages with automatic detection. Switch languages on the fly.

Global Shortcuts

System-wide keyboard shortcuts that work in any app. Customizable to your workflow.

Built for macOS

Dicta is the fastest on-device
transcription built from scratch

Native Swift and Rust. Optimized for Apple Silicon. Outperforming alternatives with hardware-aware inference.

Powered byMetalWhisperTauri

Build real-time workflows
with on-device AI

Users don't care where your model runs. They care how it feels.

Fast responses for text and audio.
Offline continuity. No network, no break.
Consistent latency. Even under load.
Multilingual

Speak your language

Dicta understands you, no matter what language you speak.

Auto-detectWhisper identifies the spoken language automatically.
Translate to EnglishSpeak any language, get English output.
100+ LanguagesMultilingual models support global voices.
EnglishSpanishFrenchGermanItalianPortugueseChineseJapaneseKoreanHindiArabicRussian+90 more
Open Source

Built in the open

Transparency and privacy aren't just features — they're our foundation. Dicta is 100% open source and always will be.

Fully Transparent

Every line of code is open for inspection.

Privacy First

Your voice never leaves your device.

Community Driven

Help shape the roadmap with us.

Customizable

Fork it, modify it, extend it.

Ready to contribute?

Check out our repository and join the community.

View on GitHub
Free Forever

Get started with Dicta

One-click install. No dependencies.
100% local processing. Your data stays private.
Global shortcuts work in any app.
Start transcribing in under a minute.
1

Download Dicta

Get the latest version for macOS

dicta-v1.0.0.dmg
2

Grant Permissions

Allow microphone and accessibility access

MicrophoneAccessibility
3

Set your shortcut

Default: Option + Space

Option+SpaceStart recording
4

Start speaking

Your voice, instantly transcribed

Press shortcut → Speak → Release → Text appears
Simple pricing

Free. Forever. No catch.

Dicta is open source and completely free to use. No subscriptions, no trials, no hidden fees.

$0/forever

Everything included. No limits.

Unlimited transcriptions
All Whisper models included
50+ languages supported
Global keyboard shortcuts
Custom vocabulary training
Voice-activated snippets
Works completely offline
No data collection
Open source (MIT License)
Community support on GitHub

Dicta is proudly open source under the MIT License.
Contributions and feedback are welcome on GitHub.

FAQ

Frequently asked questions

Everything you need to know about Dicta

Yes, completely free forever. Dicta is open source under the MIT License. There are no subscriptions, no trials, no hidden fees, and no premium tiers. Everything is included.

Yes! Dicta runs entirely on your Mac using local AI models. No internet connection is required for transcription. Your voice data never leaves your device.

Dicta supports 50+ languages through the Whisper AI model, including English, Spanish, French, German, Chinese, Japanese, Korean, Arabic, and many more.

Dicta uses OpenAI's Whisper model, which is state-of-the-art for speech recognition. Accuracy depends on audio quality and the model size you choose, but it rivals cloud-based services.

Dicta works best on Apple Silicon Macs (M1, M2, M3, M4) where it can leverage the Neural Engine for fast inference. It also works on Intel Macs, though performance may vary.

Simply press the global shortcut (Option + Space by default), speak, and release. Your transcribed text will be automatically typed wherever your cursor is. You can also customize the shortcut in settings.

Yes! Dicta supports custom vocabulary training so it can better recognize names, technical terms, and domain-specific jargon that you use frequently.

Absolutely. All processing happens locally on your Mac. No audio is ever sent to any server. No data is collected. Your voice and transcriptions stay completely private.

Still have questions?

Ask on GitHub Discussions

Free your voice.
Start transcribing locally

Download Dicta and experience AI transcription that respects your privacy. No subscriptions, no limits, no cloud required.