Can we use different better model for ocring ?

[Greatz08]

https://core.ac.uk/download/pdf/210844206.pdf

Ocr result was not great and their were many problems, so i thought if their is any better model and can we use that with docling ?

[dolfim-ibm]

This is a list of supported OCR engines in Docling: https://docling-project.github.io/docling/faq/#which-ocr-languages-are-supported.

Additionally,

1. You can add new OCR engines as a plugin, https://docling-project.github.io/docling/concepts/plugins/#ocr-factory
2. You can run the VLM pipeline which uses last-generation vision-language models which are also able to do OCR, https://docling-project.github.io/docling/usage/vision_models/