A compact, specialized model focused on a narrow task: recognizing Korean text in images. It operates quietly and efficiently, without pretense of broader capability. Its small size means it runs lean, but that same constraint limits how well it handles complex or degraded image inputs.