HunyuanOCR

Powered by Tencent Hunyuan Team

Quick Examples - Click to load

Core Features

  • Text Detection & Recognition - Multi-scene text detection and recognition
  • Document Parsing - Automatic document structure recognition
  • Information Extraction - Extract structured data from receipts and forms
  • Visual Q&A - Text-centric open-ended question answering
  • Translation - Translate text in images across 14+ languages

Usage Tips

  • Inference - For production, use VLLM for better performance
  • Image Quality - Ensure images are clear, well-lit, and not heavily skewed
  • File Size - Recommended max 10MB per image, JPG/PNG format
  • Use Cases - OCR, document digitization, receipt recognition, translation

2025 Tencent Hunyuan Team. For research and educational use.