According to Sina Technology, a new multimodal application called "Lingguang" has quietly launched on platforms such as Tencent App Store and vivo Application Store, and has started an invitation-based internal test. Users can directly log in and experience it using their phone number or Alipay account.
It is reported that "Lingguang" was developed by Alipay (Hangzhou) Digital Service Technology Co., Ltd. Its biggest highlight is the built-in "AGI Camera" feature. This feature can recognize and understand scenes and content in the real world through the camera lens, enabling "shoot and ask," real-time understanding, and answering. Industry insiders pointed out that this feature is similar to the image recognition functions of ByteDance's Doubao App and Alibaba's Yuanbao App, but Lingguang emphasizes "cognitive-level understanding" and may have stronger scene analysis and multimodal reasoning capabilities.
In fact, Ant Group has been exploring multimodal and AGI for several months. In late April this year, Ant achieved the unification of image understanding and generation for the first time; in May, it launched the Ming-Lite-omni-Preview model, which is the world's first open-source model that can rival GPT-4o in terms of modal support, possessing integrated capabilities for speech and image generation and understanding.
