FrameOCR brings cutting-edge OCR performance to macOS via vision language models without relying on external APIs or cloud processing. All OCR processing happens entirely locally on your device. Supported features include document segmentation, multi-target language support, table recognition, and LaTeX math recognition.