Tota is a Mac app that transcribes your interviews with speaker labels and lets you draft by voice, with every word processed on your Mac, so a source's recording is never uploaded to a transcription service.
You have an hour of recorded interview with someone who agreed to talk on the condition that it stays between you. The fastest transcription services are cloud services, which means the first thing you would do with a confidential recording is hand it to a company whose retention policy you have not read. Tota transcribes the file on your Mac, labels who said what, and exports a transcript with timestamps. Then it helps you write the piece by voice.
Why source protection rules out cloud transcription
Source protection is a promise that nobody other than you holds what the source said. Cloud transcription services break that promise structurally: the audio is uploaded, processed on the vendor's servers, often routed through a third-party AI provider, and kept for whatever period the terms specify. A subpoena, a breach or a change of ownership at the vendor becomes your source's problem.
Many reporters know this and transcribe confidential interviews by hand, which costs hours per recording. The alternative is a transcription tool that never uploads. An Apple Silicon Mac can run OpenAI's Whisper models locally, and on-device speaker identification can label who said what, at speeds that make hand transcription unnecessary.
Tota does both. Audio Drop transcribes files entirely on the Mac with speaker labels and timestamps, and the same app provides system-wide dictation for drafting, so notes and copy can be spoken into any editor. No account, no upload, and it works on a plane.
What journalists actually transcribe and dictate
- Recorded interviews. Dropped onto Audio Drop as MP3, M4A or WAV; transcribed with speaker labels you can rename to real names or initials.
- Press conferences and panels. Multi-speaker recordings where the diarisation does the sorting.
- Voice notes from the field. Memos recorded on an iPhone, transferred and transcribed on the Mac. See the Voice Memos guide below.
- First drafts. Dictated into Google Docs, Word, Scrivener or the CMS, then edited on screen.
- Pitches, emails and notes to editors. Short pieces of writing that are faster to say than type.
What leaves your Mac, and what doesn't
This is the whole list. Every row is checkable from the app's own behaviour, and the security page explains how to watch its network activity yourself.
| Data | Leaves your Mac? | What actually happens |
|---|---|---|
| Your voice while dictating interviews and notes | No | Recorded to a temporary file, transcribed on your Mac by Whisper via WhisperKit, then deleted. |
| The transcript | No | Typed into the app you are using. Kept in a local history for three days, then archived as Markdown in a folder you choose. Entries can be deleted. |
| Text on your screen | No | With “Spell what’s on your screen” on, the visible text of the window you are dictating into is read once, to bias spelling. Never stored or sent. Can be turned off. |
| Glossary terms and learned corrections | No | Stored as JSON files in Tota’s folder in Application Support. |
| Smart Formatting | No | Runs a local Qwen 3 language model via MLX. Off by default. |
| Audio files dropped in for transcription | No | Transcribed and speaker-labelled on your Mac. |
| Licence key and a machine identifier | Yes | Sent to api.lemonsqueezy.com at activation and validation. Never any content. |
| Update check | Yes | A download request to github.com at launch. Can be disabled in Settings. |
| Model downloads | Yes, one-way | Speech, formatting and speaker-ID model files from huggingface.co, once, on first use. |
| Usage analytics or crash reports | Never | There is no analytics or crash-reporting SDK in the app. |
How Tota works for journalists
Install the app, grant it microphone and accessibility access, and press Cmd+Shift+R wherever your cursor is. Tota records until you press it again, transcribes the audio on your Mac with OpenAI's Whisper model running through WhisperKit, and types the result into the field. There is also a hold-to-talk key for short entries and a “Hey Tota” wake word for hands-free use. It behaves like a keyboard, so it works in browser-based systems, Word, email, Notes and anything else with a text field, with no integration to request from IT. After the one-time model download it needs no internet at all.
Three optional layers sit on top. Smart Formatting runs a small local language model that fixes punctuation, turns “new paragraph” and “new list” into structure, and corrects stutters without rewriting what you meant. The auto-learning dictionary remembers the corrections you make. And Audio Drop transcribes audio files you already have, with automatic speaker labels, also on-device.
Workflows for a reporter
- The confidential interview. Record on your phone, AirDrop the file to the Mac, drop it on Audio Drop. Rename 'Speaker 1' to the source's initials and export Markdown for your notes or SRT if you need timestamps for quotes.
- Quote checking. Exports carry timestamps, so finding the exact moment in the recording to verify a quote takes seconds. Always verify against the audio before publishing.
- Writing the piece. Dictate the draft into your editor with the shortcut, or the 'Hey Tota' wake word for hands-free. Filler words are removed automatically.
- Working offline. Trains, planes and foreign hotels with untrusted Wi-Fi. Nothing in Tota needs a network after the one-time model download.
A glossary for journalists
Names are the hard part of transcribing news. A Tota glossary is a short list of terms the speech model is nudged towards while it transcribes, with sound-alike matching afterwards for anything it still mishears. A glossary for a running story might hold:
- the names of sources, officials and organisations in the story
- place names
- acronyms such as FOI, OFCOM, SEC or NHS
- publication and programme names
- technical terms specific to the beat
The built-in Names glossary is a starting point. Keep each list to the terms that get misheard; the live hint has room for roughly 70 to 90 words. Glossaries bias dictation; Audio Drop transcripts do not use them, so check names in a transcript against the audio.
Tota ships with a built-in Names starter glossary you can edit, and the glossaries guide covers set-up, the hint budget and per-app switching.
What Tota does not do
- It does not record calls or meetings. Tota transcribes recordings you already have and dictation you speak. Use your phone or a recorder, with the consent your jurisdiction requires, and transcribe the file.
- Transcripts need checking before quotation. Whisper is accurate but not perfect, and names and numbers are where it slips. The timestamps exist so you can verify against the audio.
- It does not sync or back up. Transcripts and exports are files on your Mac and nowhere else. That is the point, but it means your own backup and encryption habits are what protect them.
- Transcripts are files on your Mac. Nothing is synced or backed up by Tota. FileVault, your backup tool and your device policy are what protect them.
Pricing for a newsroom or freelance desk
Tota is a one-time purchase of £19.99, with two seats per licence, after a 14-day free trial of every feature. There is no subscription, no per-minute transcription fee and no word limit, because local processing costs us nothing per word. Larger teams buy additional licences. Pricing is the same worldwide; see the pricing section for what the licence includes.
Frequently asked questions
Is the interview recording uploaded anywhere?
No. Audio Drop transcribes the file on your Mac using OpenAI's Whisper via WhisperKit, with speaker identification also running on-device. Nothing is sent to a server, and the app works with the internet off. The only network calls Tota makes are a licence check, an update check and one-time model downloads.
How long does an hour of interview take to transcribe?
It depends on the Mac and the model you choose. Whisper on a recent Apple Silicon Mac runs much faster than real time with the default model; larger models are slower and more accurate. There are no minute limits either way.
Does it label who is speaking?
Yes. Speakers are detected automatically and colour-coded, and you can rename 'Speaker 1' to a name or initials; the whole transcript and all exports update. Exports include text, Markdown, SRT with timestamps, JSON and CSV.
Can I transcribe iPhone Voice Memos?
Yes. Transfer the memo to your Mac and drop it on Audio Drop. The step-by-step guide to transcribing Voice Memos is on the blog.
Does it do other languages?
Whisper's multilingual models transcribe many languages and can translate into English. Choose a multilingual model in Settings; the English-only models are faster if you only need English.
What does it cost?
£19.99 once, after a 14-day free trial, with two seats per licence. No subscription and no per-minute transcription fee.
Transcribe the interview without handing it to anyone
Download Tota and drop the recording on it. The transcript stays on your Mac. Two seats, £19.99 once.
14-day free trial · macOS 14 or later · No account needed
