Support

How can we help?

Answers to common questions. Still stuck? Email support@moklabs.com and we'll get back to you.

Requirements: Lavra runs on Apple Silicon — macOS 26 or later, and iOS 26 or later. The on-device intelligence (cleanup & summaries) needs an Apple Intelligence–capable device; on devices without it, you still get the transcription.

How do I transcribe an audio file?

Open Lavra and tap + Import (or drag a file onto the window on Mac). Lavra decodes, transcribes, cleans up the text, and writes a title and summary — all on your device. Supported formats include mp3, opus, ogg, m4a, wav, aac, and flac.

How do I bring in a WhatsApp or Telegram voice note?

Two ways:

  • Share (iPhone): on a voice note, tap Share and choose Lavra. The first time, tap “More” in the share sheet and enable Lavra. If you share several at once, Lavra asks whether to keep them separate or merge them.
  • Export chat (.zip): in WhatsApp, export a conversation with media, then import the .zip in Lavra. Each voice note becomes text, labeled by who sent it. (The Export Chat option is on your phone — send the .zip to your Mac to import there.)

Note: this is importing/sharing the audio you already have — not a live connection to your accounts.

How does dictation work?
  • Mac: hold the global shortcut (default ⌃⌥D) and speak; release to transcribe. You can record a custom shortcut in Settings → Dictation.
  • iPhone: assign the Action Button to “Dictate with Lavra” (Settings → Action Button → Shortcut), or use the mic button in the app. Lavra opens recording, you tap Stop, and the text is copied to your clipboard automatically — ready to paste anywhere.
Why are some words spelled wrong, and how do I fix it?

Transcription can mishear names, brands, or jargon. Add them to your Vocabulary (Settings on Mac, or the book icon on iPhone) and the AI will spell those terms exactly as you wrote them when cleaning up and summarizing. You can also edit any transcript and its title inline, like a notes app — changes save automatically.

What's the difference between the fast and high-accuracy modes?

The fast mode uses Apple's on-device speech recognition — quick and great for everyday notes. On Mac, you can switch to a high-accuracy mode (Whisper) for difficult audio; it downloads a model the first time you use it. Don't quit the app while the model is downloading.

How does syncing between my Mac and iPhone work?

If you're signed in to iCloud (with iCloud Drive and iCloud Keychain on), the text of your transcriptions syncs between your devices through your own private iCloud, end-to-end encrypted. Your audio stays on the device it was created on and is never synced. Give the key a moment to propagate the first time. If you're not on iCloud, the app still works fully and keeps everything local.

Can I merge several recordings into one summary?

Yes. Import or share multiple audios and choose Merge. Lavra combines them into a single transcript and writes one coherent summary covering all of them — not just the individual summaries stitched together.

How do I export or copy a transcript?

Open a transcription and use Copy, or export to .txt, .md (Markdown), or .srt (subtitles). On Mac, export is in the document toolbar; on iPhone, use the share button.

Is my audio sent to the cloud?

No. Transcription, cleanup, and summaries all run on your device. Lavra has no analytics, no ads, and no account. See our Privacy Policy for details.

What languages are supported?

The app interface is available in English, Portuguese (Brazil), and Spanish, and follows your system language. Transcription and summaries respond in the language of your audio.

Still need a hand?

Email support@moklabs.com — include your device and OS version and we'll help.