I’m using SBERT-style embedding models for the semantic matching, which works very well in practice.
For non-text content, the app also analyzes images (OCR + object recognition) using Apple’s Vision framework. That part is surprisingly powerful, especially on Apple Silicon.
> I need to do something for images that are already classified/tagged via FastVLM
What’s the concrete use case you’re targeting with this?
Full-time: C++ work on nearby connectivity (bluetooth) for embedded / industrial devices (factory equipment). Deep stack, hardware constraints, long lifecycles, high reliability.
Non-web work feels very different: stronger constraints, slower but deliberate releases, and bugs are much more expensive. There’s a lot of interesting software being built far away from HTTP and browsers.
Floxtop suggests the top 5 destination folders where a file could best belong. You stay fully in control: you can choose one of the suggestions, move files individually or in bulk (Move All), or select a completely custom folder location at any time.
If you change your mind, you can Undo per file or use Undo All to revert the entire operation.
It looks inside each file to see what it’s about, then moves it to the right folder for you.
Everything happens on your Mac, so nothing leaves your computer. No clouds, no servers.
It works in 50 languages (including English, German, French, Spanish, Swedish) and with images (OCR and object recognition), PDFs, Microsoft Office, ePubs, text, Markdown, and many other file types.
If you have messy folders anywhere on your Mac, Floxtop can help.
Deleted Comment
A highly social, relatively hairless bipedal ape that was once a nomadic hunter-gatherer, but has adapted to create websites. :)
1. My preferred installation method is through `brew`, would you consider adding it as and install option?
2. Any plans of supporting multiple languages?
3. Can it move files to mounted folders? I use google drive for most of my admin stuff and their client allows to mount the drive.
1. A Homebrew option isn’t available yet, but it’s something I’m open to exploring. Homebrew doesn’t currently support payments or license handling for paid apps, so Floxtop doesn’t fit perfectly into their workflow. For now, it’s available via direct download from the website.
2. Multi-language support. Yes — I’m actively working on this. Which languages would be most important for you? That helps me prioritize.
3. Moving files to mounted folders (e.g., Google Drive) Great suggestion. I’ll look into this, as supporting mounted cloud folders could be very helpful for many users.
Problem: My Downloads and Desktop folders would fill up with PDFs, screenshots, and random files, making cleanup and finding the right file frustrating and time-consuming. Solution: Floxtop uses Sentence Transformers (SBERT) to embed document text and metadata, plus OCR and object recognition to read images. It groups and classifies files by semantic similarity — all locally, so your sensitive files (finance, medical, personal, etc.) stay private.
Current version: Supports 30+ file types including images, Office documents (Word, Excel, PowerPoint), PDFs, EPUBs, Markdown, CSV, and more. Parsing and classifying 50 files takes around 17 seconds. It integrates with Finder (right-click > Organize with Floxtop), Quick Look + thumbnails, and includes keyboard shortcuts for a fast workflow. Currently, Intel Macs are not supported and it only works with English-language files.
I launched an early beta ~3 months ago, Fallinorg (https://news.ycombinator.com/item?id=44932375), and have since built out more features based on user feedback. I’d love your feedback on classification workflows, missing features, pricing, usability, and potential bulk operations or integrations. Happy to answer questions and share implementation details!
Free version here - requires macOS 15+ with Apple Silicon: https://github.com/taranntell/fallinorg/releases/download/1....
Demo video: https://floxtop.com/assets/demo.mp4