Your sessions, transcribed.
Never uploaded.
Verba turns therapy recordings into clean, speaker-labeled, supervision-ready transcripts — entirely on your machine. No account, no cloud, no audio ever leaving your device.
Windows · Works fully offline · No account
Confidentiality first
Built for the work you can’t put in the cloud.
Therapy audio is some of the most sensitive material there is. Verba never asks you to hand it to a server — because there’s no server to hand it to. Every recording, and every transcript, stays on your own machine.
ever uploaded
100% on-device
The transcription engine is built right into the app. Transcribing, separating speakers, editing, and exporting all happen on your machine — your recordings and transcripts never leave it.
No account, ever
No sign-up, no login, no cloud profile to manage. Open Verba and start working in seconds.
Works fully offline
Pull the network cable and it still transcribes. Privacy you can verify, not just a promise in a policy.
From recording to supervision-ready.
Three steps. No setup, no model downloads, no API keys — it’s all in the app.
Drop in the recording
Any common audio file becomes a named session. Queue a whole week at once and let them run.
Transcribe & separate speakers
Every turn is transcribed and split by speaker, pre-labeled 상 (therapist) and 내 (client) — swap them in a click if the order’s reversed.
Edit and export
Fix a line, split a turn, switch to the Korean supervision view, then export to txt / csv / xlsx — or a portable .vrb bundle.
오늘 한 주는 어떻게 지내셨어요?
음… 솔직히 좀 버거운 한 주였어요.
그냥 받아들이려고 하는 것 같아요.edited
그게 어떤 영향을 주는 것 같으세요?
Everything in one quiet workspace
A tool that understands a counselor’s day.
Speaker separation
Speakers are separated automatically and pre-labeled 상 (therapist) and 내 (client) by who speaks first — confirm or swap once, instead of tagging every line.
A focused editor
Inline edits, split & merge turns, undo/redo across everything, and click any word to seek the audio. Built for hours of reading.
Korean supervision format
An annotated view with numbered turns and inline (침묵 N초) silence markers — the format your supervision reports use.
Export anywhere
txt, csv, json, xlsx — or a self-contained .vrb bundle with transcript and audio together to hand to a supervisor.
The original, always kept
Verba keeps the original machine transcript and marks every segment you’ve edited, so edited text is always clear from the raw — with full undo and redo as you work.
Ready on first launch
Everything is built into the installer. No model downloads, no keys, no configuration — Verba just works the moment you open it.
Made for the way you work.
Built for supervision reports
Turn every session into a supervision-ready transcript — numbered turns, silence markers, the format your program expects. Spend your evening reflecting on the session, not typing it out.
A two-hour session, transcribed and
speaker-labeled, hands-off.
The documentation burden that used to linger after a session ends gets much lighter. Get back the evenings you spent typing it out by hand — while the audio stays exactly where it belongs.
From the clinicians using it.
(Placeholder cards — real quotes go here before launch.)
“[Placeholder] A real, short quote from an early clinician about how Verba changed their post-session workflow.”
“[Placeholder] A line about confidentiality — why keeping audio on-device mattered for their practice.”
“[Placeholder] A supervisor’s note about reviewing .vrb bundles with the audio in context.”
Start transcribing privately.
Download Verba for Windows. Your audio never leaves your machine — and never will.
Free during beta · Windows