Computer Vision Read API for Optical Character Recognition (OCR) announced the general availability of the new model with support for 164 languages. It’s also available as a Docker container.

Source