Type Write at the speed of your thoughts
Dictate in your language(s), anywhere, writing 4X faster.
Transcribe any audio. Know who said what. Get AI meeting summaries.
All 100% local on your Mac.
100% Private
Sub-200ms Latency
macOS Native
Speed you have never experienced before. Global hotkey works in any app. Hold to record, release to transcribe.
⌘⇧D
Default shortcut — fully customizable
Capture audio from Teams, Zoom, Safari, podcasts—anything playing on your Mac.
All transcription happens on-device. No cloud. No servers. HIPAA-compliant by design.
Email in Mail? Notes in Obsidian? Slack message? Yakki sees your context and adjusts automatically.
Filler words, repetition, false starts—automatically cleaned before text appears. Dictate and done.
25+ languages with instant switching. Dictate in Spanish, then English—no settings, no restarts.
Automatically identify and label speakers in your recordings.
Full searchable history with audio playback. Find anything you've ever dictated.
You have a subscription plan. Would you rather own your tools? You have a forever plan that includes all future updates.
Get key points, decisions, and action items from any meeting—automatically extracted by AI.
Drag and drop MP3, WAV, M4A, OGG, or WebM files. Transcribe recordings from anywhere.
Only on Yakki
Transcribe anything you can hear.
Other dictation apps only capture your voice. Yakki captures everything.
- Zoom and Teams meetings (undetectable, no weird participant joining your meetings)
- Podcasts and YouTube videos
- Voice memos and recorded interviews
- Any app playing audio on your Mac
Combine it with speaker diarization to get perfectly labeled transcripts of your meetings—all without sending a single byte to the cloud.

Select the app, make sure the audio is working and click record. Simple as that.
SUMMARY
In this handover meeting, James discussed updates on the testing process, including the copy to life document and event metrics. Juls provided insights on ongoing conversations and the status of various tests.
James created a personalized version of the copy to life document. 0:31
Discussion on the status of event metrics and product functionality. 0:08
James highlighted the need to check the variation code for errors. 0:28
Juls confirmed ongoing conversations regarding user tracking. 0:22
Continue monitoring the status of the library updates for reclassification.
• 0:41
To ensure the core product metrics are functioning correctly.
James to discuss analysis updates on Monday.
James• 22:11
TOPICS
testing processevent metricscopy to lifeuser trackingproduct updates
Meeting intelligence
Take perfect meeting notes
Turn your meeting recordings into clear, actionable insights.
- Summary, key insights, and next actions
- Replay the exact minute things were defined
- Identify speakers automatically
- Send the perfect follow-up email
When you choose to generate Meeting Intelligence, the transcript text is sent over an encrypted connection to our secure proxy, then to our cloud providers for analysis. Your data is never used to train AI models. You can disable cloud AI entirely with "Only Local" mode.
I love that it respects my clients' privacy. The local-first approach means I document sessions without HIPAA anxiety. Game changer.
I switched from Dragon after 10 years. Yakki is faster, feels more modern, and I can do much more with a lifetime license.
I'm really liking it. It's a huge productivity boost. An email to a client that used to take fifteen minutes just took five.
Monthly
$12/month
Cancel anytime
- Unlimited transcription
- App audio capture
- Speaker diarization
- Full history
Best - Save 30%
Annual
$99/year
~$8/month
- Unlimited transcription
- Advanced ai Enhancements
- Speaker diarization
- Full history
- A year of updates
- Save $43.89 annually
Lifetime
$149once
Never pay again
- Pay once, use forever
- Everything in annual
- Early access to new features
- Future updates
- Priority support
Yakki has strong privacy by design, but is not yet formally HIPAA certified. Audio recording, transcription, and speaker identification run 100% on-device (always). Meeting Analysis and Ai assisted features (i.e. summaries) sends text to our AI service (optional). For PHI-containing meetings today, enable 'Only Local' mode in Settings to disable cloud features entirely.
On-Device (Always)Audio recording, transcription, and speaker identification run 100% on-device.
Cloud (Optional)Meeting Analysis (summaries) sends text to our AI service.
For PHI-containing meetings: Enable "Only Local" mode in Settings to disable cloud features entirely.
No. Your audio and transcripts are never used to train our models or third-party models (like OpenAI).
No. Audio recording, transcription, and speaker identification happen entirely on your device. Only the (optional) Ai assisted features use the cloud.
Yes. You can enable "Only Local" mode to ensure no data ever leaves your device. Core features work without internet.
We don't store your audio or transcriptions. Audio and transcripts live on your Mac. Our partners for cloud processing (i.e. OpenAi if enabled) retention policy is ephemeral, any data is deleted after 30 days.
Yes. We support all data rights (access, deletion, portability). You own your data.
Only you. We have no access to your audio files or transcription content.
Ready to try the dictation app that captures everything?
7-day free trial. No credit card required.
Works on MacOS 14+ (macOS Sonoma or later)