Show HN: We benchmarked 18 LLMs on OCR (7K+ calls) – cheaper models win

(arbitrhq.ai)

5 points | by TimoKerr 7 hours ago ago

1 comments