Drag anything onto the Dock icon. DropItDown converts it, classifies it, moves it into your archive, and writes a Markdown note pointing back — so future-you can find it.
Three steps, then the app quits. Until the next drop.
Drag a file from anywhere onto the DropItDown icon in your Dock. The app wakes only for this.
Azure Content Understanding extracts the text. A language model reads it, picks a category from your existing folder structure (or invents a new one).
The original moves to Archive/<category>/. A Markdown note is written alongside, with a link back to the original. Done — app exits.
No background process. No menu bar. No watchdog daemon. The .app wakes for a drop, processes, exits.
SwiftUI binary, Apple-signed and notarized, ~95 MB on disk plus an embedded Python pipeline for the heavy lifting.
Every archived file gets a paired .md note with YAML frontmatter linking back to the original. Search your archive with your favorite editor.
Got a wrong archive? Tell it where it should have gone. The rule gets recorded so the same mistake doesn’t recur.
The archive lives in ~/Documents (or wherever you point it). Only a short content excerpt is sent for classification.
The free tier uses a hosted DeepSeek proxy. Power users plug in their own OpenAI, Claude, or DeepSeek key.
Double-click the icon to open the main window. Two tabs, that’s it.
Free · 200 classifications / month
Pay your model provider · no app fee
macOS 14 Sonoma or newer. Apple Silicon required.
Signed with Developer ID, notarized by Apple, open source on GitHub.