OCR text extractor from IMAGES

[Lex_Drobot]

hello everyone*
Im looking for a good solution for my request

my request I have many screenshots from the phone screen. They can be in English, Chinese or other languages. Even Japanese. The text direction can be either left to right or top to bottom. Sometimes there is even from right to left.

I need the OCR service to understand what language it is and what is written there.
then save it like PDF and than to GPT to operate with data .

I know about the service’s
0CodeKit
DocCrafter
Mindee
Google cloud vision
PDF_Co
ChatGPT vision

i try them, but didnt work( only chatGPT works
as i know he is not good enough . Plz help me.

the hardest sample of my images

[Lex_Drobot]

what service can give me.

• work with a large number of recognition languages
• work with images
• recognize multi-page PDFs in scan format
• recognize text vertical, horizontal, rotated pages

From image - to text
from image - to PDF (text)
from PDF (scan images) - to PDF (text) with images or without
From PDF (text) - to text
more optional options *

plz help