Pure Scribe - Speech to Text

Drop in a recording and get a clean, timestamped transcript back. Everything works offline. No account, no subscription, no upload. Ever.

Turn recordings into polished transcripts — without sending a single second of audio to the cloud.

PureScribe runs advanced speech recognition entirely on your device. Drop in an interview, a lecture, a meeting or a podcast, and get a clean, timestamped transcript you can edit, search and export. Nothing is uploaded. Nothing sits on a server. There is no account to create.

WHY ON-DEVICE MATTERS Confidential interviews, medical notes, legal recordings, unreleased material — some audio simply should not leave your hands. PureScribe processes everything locally, so your recordings stay yours. It works on a plane, in a basement, or with Wi-Fi switched off.

FAST, ACCURATE, YOURS • Pick the model that fits the job, from a lightning-fast draft to maximum accuracy • Optimised for Apple silicon • Automatic language detection, or choose the language yourself • Queue several files and let them run

A REAL TRANSCRIPT EDITOR • Readable typography built for long sessions • Tap any timestamp to jump straight to that moment in the audio • The current line highlights as playback moves • Search the whole transcript • Fix a word inline — changes save themselves

EXPORT ANYWHERE • Plain text for writing • Markdown for structured notes • SRT and WebVTT subtitles for video • JSON for developers and downstream tools

WORKS WITH YOUR FILES MP3, WAV, M4A, AAC, FLAC, MP4 and MOV — audio or video, drag and drop.

FREE AND PRO PureScribe is free to use: transcribe files up to 10 minutes, 30 minutes per day. PureScribe Pro is a one-time purchase that removes every limit and unlocks the two most accurate models, Turbo and Large. No subscription. Buy once and use it on your Mac, iPhone and iPad.

PRIVACY, PLAINLY No analytics. No telemetry. No accounts. No third-party services. The only time PureScribe touches the network is when you ask it to download a speech model.