Offline AI Transcription for Whole Folders of Audio

Most people who need to transcribe a folder of recordings don't have one file - they have years of them: old dictations, client calls, hearings, interviews, session notes, sitting untouched because transcribing them by hand never made it to the top of the list. Pro is built for exactly that backlog. Give it a folder and it works through the recordings in it one by one, in any common audio format, and writes a text document next to each original file - not into some separate library or export folder, but right beside the recording it came from, where you'd expect to find it.

It can run overnight, unattended, on a machine you already own. There's no file-size limit imposed by a server queue and no waiting on someone else's processing capacity - the only limits are your own hardware and however many hours of audio you point it at.

Recognition covers the same 90+ languages as everyday dictation, and each file's language is detected automatically, so a mixed folder - some recordings in one language, some in another - doesn't need sorting first. The document Pro writes lands beside the original recording, so the two stay together in whatever folder structure you already use to organize your archive.

Cloud transcription services work by uploading your audio to their servers, which means your recordings leave your custody the moment you use them - and you lose control over what happens to that copy afterwards. Kokukeyboard Pro never does this. Whisper, the transcription model, runs entirely on your own computer; the app has no code path that uploads audio, because there's no server on our end waiting to receive it. The only thing Pro can ever send is a bug report you choose to share.

For material where confidentiality isn't a preference but an obligation, that distinction matters: there's nothing to breach later, because nothing was ever transmitted in the first place.

Pro is aimed at people sitting on an archive of recordings rather than a single file to dictate:

In law, that's depositions, hearings, and client calls waiting to become searchable case files. In medicine, patient notes and dictations that never made it out of an audio recorder. In therapy, session recordings that need a written record without ever leaving the practice's own computer. In journalism, source interviews where keeping the recording - and the transcript - off any third-party server is part of protecting a source. In research, field recordings piling up faster than anyone has time to type them out. In insurance, claims and witness statements that need to become documents, not just audio files sitting in a folder. What connects all of it is the same thing: a backlog of audio that needs to become text, on the machine it already lives on.

Kokukeyboard Pro is $69 once - everything in the standard Lifetime license, plus batch transcription, plus a third computer, with no recurring charge. If you already own the standard license, you can upgrade for the exact difference rather than paying full price again. Every feature, including batch transcription, is included in the 14-day free trial with no card required, and purchases carry the same 30-day refund as the rest of Kokukeyboard.

Full pricing, the upgrade path, and the checkout are on the Archive Liberation page. If what you actually need is everyday dictation rather than transcribing existing files, see offline dictation for Windows; if you're weighing the pricing model itself, read why it's a one-time payment instead of a subscription.