tabi taste
teach tabi what you mean by good.
separate preference from reference. each correction gives tabi a clearer read on what to reach for next.
save what feels right. tabi turns those choices into context your agents can use.
coming soon this september
basically anything that makes you, you.
save pages, images, tabs, and selections. search by words, ideas, or the look of something.
loading current tabi interface…
tabi helps at the moment of saving, when you need something back, and eventually wherever your agents work.
1 / 7
tabi is building a scoped mcp bridge so agents can use the references you choose without getting the keys to everything.
what would be a good spinner for this component?
discombobulating...
one personal library. useful context for every compatible agent.
10 each month
or $96 for a year
your first month starts when access opens. no card required.
when smart captures run out, basic saves keep going.
coming soon this september
back to the window