Promote Sign in Sign up

Multimodal model

LightOnOCR 2

LightOnOCR-2-1B is LightOn’s flagship 1B-parameter OCR vision-language model for turning PDFs, scans, and images into clean, naturally ordered text. It is designed as an end-to-end OCR system that handles tables, forms, receipts, multi-column layouts, and math without relying on brittle external pipelines.

LightOn Multimodal Open weights apache-2.0
Visit model website ↗ Hugging Face
Type
Multimodal
Weights
Open
Downloads
212k
Likes
834
Pipeline
image text to text
License
apache-2.0

Model hub

lightonai/LightOnOCR-2-1B

Repo
lightonai/LightOnOCR-2-1B
Library
transformers
Access
Public
Technical tags and languages
en fr de es it nl pt sv da zh ja transformers safetensors mistral3 text-generation ocr document-understanding vision-language pdf tables forms image-text-to-text conversational
Open on Hugging Face

Papers

2

More from LightOn

3

About LightOn

1

We are pioneers in Generative AI, building secure and customizable solutions for enterprises to improve productivity and competitiveness.