Smart Jelly is a macOS desktop AI assistant that combines dictation, chat and voice Q&A, meeting prep and notes, to-do tracking, document drafting and web agents on top of one local memory of your work.
Smart Jelly is a macOS desktop AI assistant that combines dictation, chat and voice Q&A, meeting prep and notes, to-do tracking, document drafting and web agents on top of one local memory of your work.
More detail
It drafts emails, documents and slides but does not send or buy anything without an explicit button press, and no bot joins your calls.
Memory is stored in an encrypted database on your Mac, while reasoning is handled by third-party AI and speech providers.
Best for
Individual Mac users on Apple silicon (macOS 14+) who work across email, chat, calendars and calls and want recall, dictation and drafting in one app with a single subscription.
More detail
Fits people who want their work memory kept locally and no meeting bot in the participant list.
Pricing
One subscription covering chat, dictation, voice, meetings, documents and agents, with plans differing only by monthly feature allowances.
More detail
Basic is free with small monthly limits (2 hours of meeting notes, 20 minutes of dictation, 20 minutes of live conversation, 50 Ask Jelly answers, 2 documents/de
Main limitation
Windows, Linux, iOS or Android users, since no non-macOS version is listed.
More detail
Also poor for anyone who needs a system of record or independently validated enterprise software: the vendor labels the product a private beta and discloses that it expects defects and that it should not be relied on as a system of record.
Smart Jelly is a macOS desktop AI assistant that combines dictation, chat and voice Q&A, meeting prep and notes, to-do tracking, document drafting and web agents on top of one local memory of your work.
It drafts emails, documents and slides but does not send or buy anything without an explicit button press, and no bot joins your calls.
Memory is stored in an encrypted database on your Mac, while reasoning is handled by third-party AI and speech providers.