Extract data from documents into JSON
Count tokens in your text
Reasoning model specialized for OCR/Markdown generation.