←News detail

ModelBest OpenBMBUpdated Oct 30 14:24

MiniCPM-OCR 1.0 locally processes 415-page S-1 filing

AI summary

ModelBest OpenBMB thanks atomic.chat for the demo: running the 1.2GB MiniCPM-OCR 1.0 locally on an M5 Max MacBook, it processed all 415 pages of Anthropic's S-1 filing at about 4 seconds per page, with data never leaving the device throughout.

Quoted post

atomic.chat

@atomic_chat_hq

A 1.2GB model on a Mac read Anthropic's whole IPO 📄 @OpenBMB MiniCPM-OCR 1.0 went through all 415 pages of Anthropic's S-1 on an M5 Max MacBook at about 4 seconds a page and nothing ever left the laptop GGUFs -> https://t.co/HXaqK2YdR6 https://t.co/LuNjYUSOAv

View original

A 1.2GB model on a Mac read Anthropic's whole IPO 📄 @OpenBMB MiniCPM-OCR 1.0 went through all 415 pages of Anthropic's S-1 on an M5 Max MacBook at about 4 seconds a page and nothing ever left the laptop GGUFs -> https://t.co/HXaqK2YdR6 https://t.co/LuNjYUSOAv

View original post

Body

Huge thanks to @atomic_chat_hq for the awesome demo! 🙌

They ran MiniCPM-OCR 1.0 locally on an M5 Max MacBook to process all 415 pages of Anthropic’s S-1 filing — with a 1.2GB model, ~4 seconds per page, and zero data leaving the device.

Small model, serious document understanding, fully local. 🔥

Source

ModelBest OpenBMB