Coming soon for Mac

Turn anything you can hear on your Mac into searchable text.

Transcribbler makes accurate, speaker-labelled transcripts of your recordings, meetings and calls — and does it entirely on your own machine.

Your audio never leaves your Mac. Transcription runs on-device, on Apple Silicon. Nothing is uploaded.

{{ heroNote }}

Your Mac ON-DEVICE
audio stays inside · never uploaded
MicSystem audioWhisper · on-device
Native Mac app · Swift & SwiftUI Requires macOS 14.4 or later Apple Silicon Sandboxed In active development

Privacy

Your audio never leaves your Mac.

Most transcription apps upload your recording to a server you don't control. Transcribbler doesn't. Every transcript is made on-device, using the Apple Silicon in your Mac. There's no account to make, and your recordings are never sent to us — because there's nowhere for them to go.

On-device by default

Recording, transcription and speaker detection all happen locally. No upload step, ever.

Stored on your Mac

Your conversations live in your library on your machine — not in someone else's cloud.

AI is opt-in only

The optional AI features are off until you turn them on, and even then your audio is never sent — only transcript text, to a provider you choose.

Cloud transcription tools upload your conversations, then ask you to trust their delete button. Transcribbler never uploads them in the first place.

This matters most when what you're recording is personal — a doctor's appointment, a therapy session, a difficult call. It should stay yours.

Capture

Get audio in, any way you have it.

Files, your microphone, whatever's playing on your Mac, or a folder that transcribes itself. However the sound reaches you, Transcribbler can take it from there.

Drop in files

Audio or video — mp3, wav, m4a, mp4, mov. Drag in a batch and transcribe them all at once.

Record your microphone

Hit record and capture straight from your mic, right inside the app.

Record system audio

Capture whatever is playing on your Mac. macOS asks for Screen Recording permission once — that's the only way any app can reach system audio. Transcribbler takes the audio only, never a picture of your screen.

Meeting mode

Record your mic and the system audio together, mixed into one recording — both sides of the meeting.

Watch folder

Point Transcribbler at a folder. Anything new that lands there transcribes itself, hands-off.

Menu-bar quick record

A fast voice note from the menu bar that files itself away in your library.

System-wide dictation

Press ⌥⌘D in any app, speak, and the text lands on your clipboard.

Live Captions

A floating, always-on-top window that transcribes in realtime as you speak or listen.

Automatic call detection

When a call starts in Zoom, FaceTime or Microsoft Teams, a small corner prompt asks whether to record. Always dismissible — and you can switch it off entirely.

Accuracy

Choose the engine that fits the job.

Different recordings want different trade-offs. Transcribbler ships several on-device engines and lets you pick — the fastest, or the most careful — with tools to nudge the result the rest of the way.

The Voice Engine screen: Whisper shown as the active on-device model with speed and accuracy readouts, and other selectable models including Parakeet TDT English and multilingual.

Multiple on-device engines

OpenAI Whisper (Tiny through Large v3), NVIDIA Parakeet TDT — an English model and a multilingual one — and Apple's built-in speech.

Fastest, or most accurate

Pick the trade-off you want. A quick draft or a careful pass — it's your call, recording to recording.

Custom dictionary

Teach it the names, medications and jargon it keeps getting wrong, so it stops getting them wrong.

Hallucination filter

Silently strips the phantom text Whisper tends to invent during long stretches of silence.

Per-segment confidence

See which lines the engine was unsure about, so you know exactly where to double-check.

Translate to English

Transcribe audio in another language and translate it into English in the same pass.

Speakers & library

Know who said what — and find it again later.

Transcribbler separates the voices in a recording and keeps every conversation in one searchable place, organised by the people in them.

Speakers

Automatic speaker detection

On-device diarization labels each voice Speaker 1…N as it goes — no cloud, no setup.

Voice profiles

Enrol someone's voice once and they're recognised by name in every recording after.

Rename, reassign, merge

Tidy up speakers by hand, and see per-speaker talk-time on a colour bar — so you can tell who dominated the conversation.

Working with a transcript

Click to jump

Click any line to jump to that moment in the audio, with a waveform scrubber to move around by ear.

Playback, your way

Variable speed from 0.5× to 3×, loop a segment, and skip silence automatically.

Find, replace, star

Find and replace across the whole transcript, and star the lines that matter.

Filler-word filter & compact mode

Hide the ums and ahs, and tighten the layout when you just want to read.

Word cloud

A quick picture of what a conversation was actually about.

Merge recordings

Stitch several recordings into one continuous transcript. Non-destructive — the originals stay put.

Your conversation library

Every conversation in one place

A dashboard with title, participants, duration, tags and a snippet of each one.

Filter, sort, search

By person, tag, date, length, engine or language — plus full-text search across everything.

A view per person

Click someone to see every conversation you've had, how long you've spent talking, and when you last spoke.

Titles & tags, written for you

Auto-generated titles and topic tags, so the library stays legible without busywork.

The conversation library: a scrollable list of conversations, each with a title, participant chips, duration, word count, tags, and a snippet, with filter and search controls above.

AI features

Optional. Off by default. Your own key.

AI is entirely opt-in, and you bring your own Anthropic or OpenAI key. Your audio is never sent anywhere. When you switch these on, only the transcript text goes to the provider you chose — nothing more.

Summary

An overview, the key points, action items — with who committed to them — and the decisions that were made.

Ask

Put a question to a conversation and get an answer with timestamp citations, so you can jump straight to the line it came from and check it yourself. It's a language model reading your transcript — it can still get things wrong, which is exactly why every answer is cited.

Discover

Search by meaning across every conversation. Search “money worries” and it finds “can't afford it” and “bills piling up”. Only the query leaves your Mac; the transcripts never do.

Redaction, consent & control

Careful with sensitive recordings, by design.

Redaction mode

Automatically detects names, phone numbers, emails, addresses, card numbers and social security numbers. Redacting the transcript is reversible — the original text is never destroyed — and redacted terms are left out of exports, AI prompts and search. You can also redact the recording itself, which rewrites the audio file and cannot be undone; Transcribbler says so before it does it.

Two-party-consent reminder

Set your state and Transcribbler warns you before recording where the law requires everyone's consent. It's explicitly not legal advice, and you can switch it off entirely, whatever your state.

Stored locally

Everything — recordings, transcripts, voice profiles — is kept on your Mac. It's yours to move, back up or delete.

Export

Take your transcripts anywhere.

Export in whatever format the next tool needs — one conversation or many at once — and shape the output exactly how you like.

.txt .srt .vtt .csv .md .rtf .pdf

Bulk export

Send out many conversations at once, in the format you need.

Template editor

Design your own export format — header, body and footer sections with tokens — so it drops straight into your workflow.

About

Made by one person, on purpose.

I'm Maddie, and I make Transcribbler entirely on my own — one autistic developer who cares, maybe a little too much, about getting the small things right.

Building software like this is how I earn my living. Traditional employment has never sat well with the way my mind is wired, so I've built something that does: apps I believe in, made for real people, and supported directly by the people who use them. Every purchase and every bit of support comes straight to me — no investors, no company taking a cut — and it's what makes it possible for me to keep creating.

So, from me to you: thank you. You're not just buying an app — you're making this whole way of living possible.

Maddie's handwritten signature

Maddie

maker of Transcribbler

Coming soon for Mac

Be the first to know when it's ready.

No newsletter, no drip campaign. One email, once — the day Transcribbler is available.

{{ closeNote }}