Tinta Voice for macOS

Offline dictation and file transcription for Mac

Hold a hotkey and say what you want to write. Tinta turns it into text in your Mac apps. Have a recording instead? Transcribe the file and export the text. Speech recognition runs on your Mac.

Speech stays on your Mac One-time price Dictation and audio files

Runs locally on Apple Silicon. 3-day free trial. One-time license fee.

Download 3-day trial Buy for $19.99

Launch promo: $19.99 $24.99. 14-day money-back guarantee.

Notes — Dictation Input

// Hold hotkey to write...

Hold ⇧Right Shift to speak

Recording

Write by speaking. Keep the processing on your Mac.

Tinta is a lightweight native macOS utility that turns your speech into text without sending recordings to a cloud transcription service or charging a recurring fee.

Your recordings stay local

Speech recognition runs locally on your Apple Silicon Mac. Your voice recordings and transcript text are not sent to a cloud transcription service.

No recurring subscriptions

One payment covers Tinta Voice and its future updates. There is no monthly transcription bill or per-minute fee.

Works where you write

Press your hotkey and dictate into the active text field. If an app blocks automatic insertion, Tinta shows a copy card so the transcript is not silently lost.

System dictation in three steps

Rather than opening a separate browser tab or transcription app, Tinta is a system utility that works natively where you write. The same floating pill you see above walks through three states — here they are side by side.

Recording

01 · Hold hotkey

Transcribing

02 · Release hotkey

Done

03 · Text appears

Hold hotkey

Press and hold ⇧Right Shift — Tinta starts listening locally.

Speak naturally

Talk as you would to a colleague while Tinta records locally. When you release the hotkey, the local model transcribes your speech.

Release to apply

Release the hotkey to transcribe. Tinta inserts the text where you are writing, or gives you a copy card if the app blocks insertion.

Turn an audio or video file into text

Drop a recording into Tinta, choose the language, and transcribe it on your Mac. Save the result as TXT or Markdown, or export SRT subtitles for your video editor. See how local file transcription works. The animation below shows the steps; transcription happens in the Mac app.

Drop an audio or video file

…or wait — the demo will run in a moment.

Meeting & Call Transcripts

Drop a Zoom, Meet, or Teams recording to create a local transcript with optional timestamps.

Lectures & Seminars

Turn hours of educational lectures or seminar recordings into local transcripts you can save as TXT, Markdown, or SRT.

YouTube Video Analysis

Extract speech and dialogue text from downloaded video recordings or audio tracks without third-party services.

Start with the trial, then decide

Tinta Voice License

$19.99 $24.99

Current launch price. One-time license fee.

Push-to-talk system dictation
Unlimited offline file transcription
Markdown transcripts with timestamps
Native Apple Silicon acceleration
All future Tinta Voice updates
14-day money-back guarantee

Download trial Buy now

Frequently Asked Questions

Can I try Tinta Voice before buying?

Yes. The trial is fully featured for 3 days, requires no account, and doesn't ask for a credit card.

What are the system requirements?

macOS 14 Sonoma or later and an Apple Silicon Mac (M1, M2, M3, M4, or newer) are required for local machine learning acceleration.

Does Tinta Voice need an internet connection?

Internet is needed to download the speech model and activate a purchased license. Speech recognition then runs locally. In the current release, a paid license also needs an online check at least once every 30 days.

How does Tinta protect my privacy?

Tinta does not send dictation audio or transcript text to cloud transcription APIs. After initial model setup, speech-to-text runs locally on your Mac. Tinta contains no tracking or telemetry of what you say.

How accurate is the transcription?

Tinta Voice uses a local Whisper speech-recognition model. Accuracy varies with language, microphone quality, background noise, and speech clarity. The full trial lets you test it with your own voice and audio files.

Does it support multiple languages?

Yes. You can select Chinese, Danish, Dutch, English, French, German, Italian, Japanese, Korean, Norwegian, Polish, Portuguese, Russian, Spanish, Swedish, and Turkish. Automatic language detection is also available. The app interface is in English. Try your own voice and recordings during the free trial; accuracy varies by language and audio quality.

Does Tinta Voice rewrite or modify my words?

No. Tinta transcribes speech without an AI rewriting layer. Speech recognition can make mistakes, so review important text before using it.

What happens if my cursor is not in a text field? Will I lose my transcription?

No — Tinta never silently drops your text. When the active focus isn't an editable field (or auto-paste is blocked by an app), a small floating card slides in with your transcription, a Copy button, and a 15-second countdown ring. Click Copy and paste with ⌘V wherever you want. If you do nothing, the card simply fades — no clipboard pollution, no data leaving your Mac.

What is your refund policy?

If Tinta Voice doesn't work for you, request a refund within 14 days through the refund link in your Lemon Squeezy receipt — no questions asked. All payments and refunds are processed by Lemon Squeezy, our Merchant of Record. For setup or technical questions, email [email protected].