A focused document reader that takes images of text and converts them to machine-readable output. It handles the unglamorous but essential work of extracting text from scanned pages, photos, and documents across multiple languages and scripts. Its open-weight nature means it can run locally without sending documents to external services.