π LightOnOCR-3 is hereπ LightOnOCR-3 is here. Our most accurate document understanding model.
Coming soon.