Private, on-device interview transcription

On the record.Off the server.

Kayit transcribes your interviews locally on your Mac, with speaker identification and quote management built in — so your audio, and the transcription, never touch a server.

Why local matters ↓
Mac with Apple chip (M1 or later)
No account required
Free trial
source-interview-04.m4a● On device
00:00:08ReporterWhen did you first become aware of the situation?
00:00:17Maria ChenBack in March, long before anyone was talking about it. I tried to report it then but was told to wait.
00:00:41ReporterDo you have documentation?
0:41 / 3:12
Insightson device
Summary
Source first flagged the issue in March but was told to wait before anyone acted on it.
Outline
00:17 The March warning
00:41 Documentation & records
Entities
People
Maria Chen4Tom Reyes3Councilman Ortiz2
Organizations
City Council3Public Works2Budget Office2
Places
City Hall3Riverside District2Fifth Ward
Dates
March 20252January 12April 3
Key Terms
budget5documentation4inspection3shortfall3permit2
Where your audio goes
Descript, Otter.ai, Whisper API, Rev, Trint — all upload your audio to the cloud
Kayit — stays on your Mac
01

Picks the right model for the job

Choose between Apple's SpeechAnalyzer, Parakeet, or Whisper depending on accuracy, speed, or language needs — swap per interview. All run on-device; none of them are a network call.

02

Knows who's speaking

Automatic speaker identification separates voices without cloud tools. Rename speakers in your notes after the fact.

03

Insights, computed on-device

An optional on-device AI gives you a plain-language summary, a click-through timestamped outline and suggested pull-quotes — plus the people, organizations and places mentioned. It all runs on your Mac, and switches off per interview or across your whole organization depending upon your generative AI policy.

04

Built for your workflow

Audio and text stay in sync — click any word to jump to that moment, or save any passage as a quote. Export transcripts as text, Markdown, CSV or subtitles, save quotes and audio clips, plus an open .transcript.json sidecar that drops straight into Obsidian. Your files, no lock-in.

Who it's for

Your work runs on trust.
Kayit is built to keep it.

A confidential source's recording is the single file with the most to lose if it leaks. A cloud vendor's breach, a subpoena, a quiet change to their terms of service — none of that is a risk you can fully assess, because none of it is in your hands.Kayit's answer isn't a stronger promise. It's removing the transfer entirely: the audio never leaves the machine it was recorded on, so there's no vendor, no server, and no third party to compel, breach, or subpoena.

Built first for journalists, but useful to anyone bound to a confidence:Lawyers (privilege) · Researchers (consent & IRB) · Clinicians (patient confidentiality) · Investigators (sensitive casework). If your IT or security team has blocked cloud transcription outright, Kayit is the tool they can actually approve.

  • Lawyers privilege
  • Researchers consent & IRB
  • Clinicians patient confidentiality
  • Investigators sensitive casework
Questions? Read the FAQ ↓
  • No account needed. Nothing to sign into and no personal data to hand over.
  • Works offline. After a one-time model download, disconnect the Wi-Fi and it still runs.
  • You own the files. Export transcripts, quotes and audio clips as Markdown, plain text or an open .transcript.json sidecar that drops straight into an Obsidian vault — no lock-in.
  • Cloud-folder warnings. Kayit warns you before saving or exporting interviews into iCloud, Dropbox or other syncing folders.

Pricing

One simple plan. Everything on your Mac.

Start with a free trial — no account, no per-minute fees and no upload limits. Cloud tools meter your minutes and cap your uploads; Kayit doesn't, because your audio never leaves your machine.

Monthly
$9.99/ month
Billed monthly. Cancel anytime.
Free trial at launch · then $9.99/mo or $99/yr · MacOS 26+ (Tahoe) · Mac (M1 or later)