OCRBench-V2 (zh)

OCRBench v2 Chinese subset: Enhanced benchmark for evaluating Large Multimodal Models on visual text localization and reasoning with Chinese text content

id ocrbench-v2-(zh) · max 1 · 11 models reported