JesRecap for professionals
If you bill by the hour and your calls are covered by a confidentiality obligation, a cloud transcription service is a problem before it is a convenience. JesRecap records and transcribes on your own machine instead.
There is no separate Pro tier
JesRecap has one version and one price. Nothing on this page is an upsell — dual-track capture, local transcription, speaker labels, meeting actions and every export format are in the €34.95 license. This page is simply about why professionals tend to be the people who need all of it.
Written for the calls you cannot upload
Different work, same two constraints: the conversation is the deliverable, and it is not yours to share.
Consultants and agencies
Client calls turn into scopes, decisions and invoices. Having the exact wording of a commitment three weeks later is worth more than a tidy summary.
Advisors and legal work
Privileged and confidential conversations do not belong in a third party's storage. On-device recording keeps the question of who else has a copy from arising at all.
Clinicians and therapists
Patient conversations sit under rules that make cloud transcription hard to approve. Processing on the device you already control is designed to support your compliance process rather than complicate it.
Researchers and interviewers
Participants consented to talk to you, not to a vendor. Both audio tracks and the transcript stay local, and you can retranscribe from the originals when your analysis changes.
Recruiters
Candidate calls contain personal data you are accountable for. Fewer copies in fewer places is the simplest way to stay accountable for it.
Anyone under an NDA
Plenty of contracts prohibit sending customer conversations to subprocessors. JesRecap has no subprocessor for your meeting content because it never transmits it.
What you actually get
Both sides, on separate tracks
Your microphone and the meeting audio are recorded as two files, then mixed into a playback track afterwards. macOS uses Core Audio process taps, Windows uses WASAPI loopback.
Transcripts produced locally
Whisper-family models run on your processor. The model is downloaded once and reused, and a live transcript during the call is optional.
Who said what
Speaker labels are applied after recording by a local diarization model — a one-time 54 MB download — separating you from the other voices.
Meeting actions
One click gives you a summary, action items, or a prompt you wrote. Either it copies a ready-made prompt with the transcript included, to paste into whichever AI you already pay for, or — with the optional AI notes setting on — a model on your own machine writes the note for you. No cloud AI, and nothing sent anywhere.
Exports that fit your files
Markdown notes for your knowledge base, plain text, a JSON archive, SRT and VTT captions, and the WAV audio — mixed, microphone only, or the other side only.
A library you can search
Meetings live in a local SQLite database beside the audio. Because both source tracks are kept, any old meeting can be transcribed again with a different model.
The privacy model, spelled out
Not a promise about our intentions — a description of what the software does and does not transmit.
| Data | Where it goes |
|---|---|
| Meeting audio (both tracks and the mix) | Stays on your device |
| Transcripts | Stay on your device |
| Speaker labels and meeting metadata | Stay on your device |
| Speech models | Downloaded once from models.jesrecap.com |
| License activation | Checked against our license service |
| Accounts and sign-ins | None exist |
| Telemetry and analytics | None collected |
| Meeting bots and participants added to the call | None, ever |
Because processing happens on the device, JesRecap is designed to support your compliance process where cloud transcription is hard to approve. That is an architectural statement, not a certification — read security and the GDPR guide before you commit.
How a billable call goes
Press record, then work
No bot joins, no link to paste, no participant appears. Tell the other side you are recording and get on with the conversation.
Let it transcribe
When you stop, the tracks are mixed and transcribed on your machine, then speaker labels are applied.
Turn it into your deliverable
Generate a summary or action items on the device, copy the prompt into your own AI, or export Markdown straight into your notes and time records.
Questions professionals ask
Is there a Pro version I should be buying instead?
No. There is one JesRecap with every feature in it — licenses start at €34.95 for one device, with 2, 4, and 8-device tiers. The only other option is volume licensing when a whole team needs it.
Can I claim JesRecap does not process client data off-site?
Audio, transcripts and metadata stay on the device — that is how the app is built. What JesRecap does contact is the model host for a one-time download and the license service for activation. Neither carries meeting content. Put that in front of whoever reviews your vendors and let them judge.
Does it summarise the meeting for me?
It can, if you ask it to. Switch on "AI notes on this device" and a local model writes the summary or action items under the buttons, saved with the meeting — a one-time download of about 1.3 GB, then no network at all. Leave it off, the default, and JesRecap copies a well-formed prompt with the transcript included so you can paste it into the assistant you already use and trust. Either way the data-handling decision stays yours, and there is no cloud AI involved.
What about in-person meetings?
They work through the microphone alone. There is no system audio to capture, so you get a single-source recording, transcript and labels.
Can I keep the raw audio for my records?
Yes. Both original tracks are kept alongside the mix, and all three can be exported as WAV. It also means you can retranscribe an old call later with a better model.
Do I need consent from the other party?
Recording rules vary by jurisdiction and by profession, and JesRecap deliberately does not announce itself in the call. Getting consent is your call to make, and we would strongly encourage you to make it out loud.
Is it ready for daily client work?
It is a preview. The full path — record, transcribe, label, export — works, but defaults and interfaces still change between builds. The releases page is the honest picture.
Your client conversations, on your hardware
One license, from €34.95 once. Nothing leaves the machine that recorded it.