Store notes, documents, and media. Find them by meaning, not keywords. Keep Notes uses semantic search to surface what matters — when it matters.
Search by what you meant, not what you typed. Semantic similarity finds related content even when the words don't match.
Structured tags for catalog queries. Combine with semantic search for both "find related" and "show me all" in one system.
Audio, images, documents — not just text. Auto-extracts metadata from files. Summarizes content so you can search everything.
Every change is tracked. See how your thinking evolved. Roll back when you need to.
Clean REST API that maps to simple operations: put, get, find, list. Build on top of it or use the CLI directly.
Your data is yours. We handle the infrastructure; you own the content.
$ keep put "Meeting with Sarah about the Q3 roadmap. Decided to prioritize auth refactor over new features." stored → %a3f7c921 $ keep find "what did we decide about priorities" %a3f7c921 (0.82) Meeting with Sarah about the Q3 roadmap. Decided to prioritize auth refactor over new features. $ keep put file://recording.mp3 -t meeting=standup stored → %b8e2d445 (transcribed, 3m42s)
Three layers, one system. Store anything, find everything.
Put in text, files, audio, images. Content is embedded and indexed automatically. Metadata is extracted from media.
Tags create hard links between items. Semantic embeddings create soft links. Both are searchable, both are useful.
Define rules that automatically bring relevant items to your attention. Untagged items, duplicates, open threads — surfaced, not buried.
Try free for 30 days. Keep building from there.
We're launching soon. Join the waitlist and be first in line.