Yeah, I’ve been in the same situation for a while, and honestly, it can get frustrating when an OCR solution works fine for Latin alphabets but just totally fails on other scripts like Hebrew or CJK characters. Recently, I experimented with OCR Studio, which claims to support over 100 languages including Latin, Cyrillic, Arabic, Farsi, Hebrew, Hindi, and even the whole CJKV group. What really surprised me was how accurately it could pick up characters even from slightly damaged or skewed documents, something that other tools I tried just couldn’t manage. Another cool thing is that it provides confidence scores for each field it recognizes, so you know exactly how reliable the reading is before committing the data. If you’re interested, they have a lot of info and demos on their site here: https://ocrstudio.ai/. It’s not just about the number of languages; it’s also how they handle multi-script documents without needing multiple passes or manual adjustments, which saves a ton of time if you’re processing a lot of IDs or forms from different countries.