Tencent has rolled out a significant update to its HunYuan multimodal understanding model, now available on the Tencent Yuanbao platform. This enhancement allows Yuanbao to process up to 10 images simultaneously, moving beyond its previous single-image processing capability.

The HunYuan multimodal understanding model integrates various information formats such as visual, textual, and layout data, providing a deeper comprehension of the relationships between elements in an image. This improvement supplies language models with richer foundational data for reasoning tasks.

Users can now upload a maximum of 10 images at once on the Yuanbao platform. This multi-image upload feature is supported across mobile, desktop, and web versions. Mobile users need version 2.11.0 or higher to select multiple images, while desktop users require version 1.8.0 or above for drag-and-drop uploads or screenshot functionality via shortcuts. The web version also fully supports this feature.

This new capability significantly boosts efficiency and user experience when handling multiple images. In scenarios that involve structuring content, extracting key points, or generating materials, Yuanbao delivers more comprehensive and precise understanding and responses.