On April 25th, Quark AI Super Box, a product of Alibaba, announced a groundbreaking new feature: "Photo Ask Quark." This innovative function, powered by advanced visual understanding and reasoning models, provides more accurate answers to a wide range of questions users encounter in the physical world.

In daily life, users often encounter situations difficult to describe accurately in words, especially when dealing with complex objects, tables, and graphics. Quark's "Photo Ask Quark" function precisely identifies people, objects, and details within images, deeply understands user needs, anticipates related questions, and thus offers more precise answers.

Quark AI Super Box launches new 'Photo Ask Quark' feature: Just take a photo, get any answer

For example, when users photograph artifacts, the function automatically links the historical context; when uploading product images, it can directly jump to the Taobao link for the same item. Whether it's people, animals, plants, buildings, landscapes, food, artwork, products, foreign language materials, or error codes, Quark can identify them within seconds.

Furthermore, "Photo Ask Quark" integrates search, scanning, image editing, translation, and creation capabilities into a single, unified interface. For complex images, users can upload up to 10 images at once for in-depth reasoning and complex task completion.

Quark currently boasts over 100 professional Agents and atomic capabilities, covering areas such as search, health, learning, travel, products, and creation. While traveling, users can utilize "Photo Ask Quark" to obtain artifact explanations and sightseeing guides; at work, it can assist in analyzing data charts, optimizing code logic, and generating work summary documents; in the health sector, uploading medical reports allows it to provide abnormal indicator analysis and rehabilitation suggestions.

In entertainment scenarios, "Photo Ask Quark" supports fun features like game guidance, screenshot-based drama identification, and couple profile picture matching, and can easily perform high-definition conversion, watermark removal, and original source tracing. Additionally, the function supports questions and interpretations in multiple languages.