Jot Cloud consent and switching back
Jot Cloud is an opt-in cloud engine. First use shows a consent dialog titled Jot Cloud sends audio to the cloud. Cancel keeps the previous engine. You can switch back anytime from Settings → Model → Engine or the menu-bar picker.
What does the Jot Cloud consent dialog say?
The first-use dialog is titled Jot Cloud sends audio to the cloud. While selected, each dictation's audio — including saved recordings you re-transcribe — is sent over an encrypted connection to whisperjot.com, transcribed, and returned. Audio is never stored on our servers. Local engines stay on-device. You can switch back anytime. Jot Cloud requires internet and an active subscription.
Cancel keeps the previous engine. Consent is remembered, so the dialog does not repeat every session. Jot Cloud is opt-in and runs only while that engine is selected.
What are Jot Cloud's limits, and how do I switch back?
There is one upload per dictation at the end — no live partials. Recordings over about 10 minutes are refused client-side. After three consecutive failures the app treats the service as down for 10 minutes and fails fast; your audio is still saved by crash recovery. WhisperJot never auto-switches to a local engine, because that could trigger a large download.
Switch back in Settings → Model → Engine, or from the menu-bar engine picker. The Privacy pane shows On — Jot Cloud selected or Off — using local engines. The opt-in Jot Cloud engine is never used for the wake word or for Meetings.
Straight answers.
What happens if I cancel the Jot Cloud consent dialog?
Cancel keeps the previous engine, so you stay on Jot Local, Jot Local Pro, or Auto. Jot Cloud is opt-in; nothing is selected until you accept. Consent is remembered after you accept once. You still need internet and an active subscription whenever Cloud is selected. Local engines stay on-device. You can open the engine picker later if you change your mind.
Why doesn't Jot Cloud show live partial text?
Jot Cloud uploads once at the end of the dictation, so there are no live partials. Local engines can show words while you speak. Recordings longer than about 10 minutes are refused on the device before upload. After three consecutive failures, the opt-in Jot Cloud engine is treated as down for 10 minutes and fails fast; crash recovery still keeps the audio.
Will WhisperJot fall back to Jot Local if Cloud is down?
No. After three consecutive failures it treats the service as down for 10 minutes and fails fast instead of auto-switching to a local engine, which could start a large model download. In-flight audio is still saved by crash recovery. Switch to Jot Local, Jot Local Pro, or Auto yourself from Settings → Model → Engine. The opt-in Jot Cloud engine is never used for meetings.
Updated 2026-09-01
Keep reading.
What WhisperJot keeps on your device
With Jot Local or Jot Local Pro, transcription runs on your computer by default and history stays in a local database. The …
Read article →Choose a WhisperJot speech engine
WhisperJot offers Jot Local for short utterances, Jot Local Pro for long-form accuracy, opt-in Jot Cloud, and Auto, which picks Local or …
Read article →Why a WhisperJot meeting will not start
Meetings always transcribe on-device. If the opt-in Jot Cloud engine is selected, WhisperJot falls back to a cached local model. With no …
Read article →Ready to dictate?
Download WhisperJot for Mac, Windows, or Linux and start with a hotkey.