Thanks for the feedback! We're working a lot more on speed in the coming few weeks :) More GPUs and more optimizations. Our has been focus on quality of outpu

Thanks for the feedback! We're working a lot more on speed in the coming few weeks :) More GPUs and more optimizations. Our has been focus on quality of output first and we'll make optimizations as we grow :) The lite models are great for simple use cases but won't don well in more complex OCR use cases.