Code release

[langerapp]

Hi Is the code going to be released anytime soon? Maybe I can you with finishing it?

Cheers

[floriancochard]

@langerapp not sure how relevant the code is today versus when I wrote it in 2019. Most recent LLMs (incl. Google Gemini, Langchain) support inference over multimodal inputs (e.g., images) with complex layouts, diagrams, or scans. They are more likely to give better results than current Tesserract implementation. Still, I've refactored it and added some docs.