On November 25, Tencent Hunyuan unveiled its open-source OCR model, HunyuanOCR, which stands out with a mere 1 billion parameters. Built on a native multimodal framework, this model has achieved state-of-the-art (SOTA) performance across a range of OCR application evaluations. It is equipped to handle translations in 14 minority languages, making it ideal for use in scenarios such as multilingual document processing, invoice data extraction, video subtitle recognition, and photo-based translation.
