The breakthrough of generating a dynamic world that can be played and even "cast spells and battle" with just one click on a mobile phone was first achieved by the Lingguang App.

On July 9th, the Lingguang App officially upgraded its World Model Experience feature. The new version integrates the latest world model LingBot World 2.0 released by Ant Lingbo, allowing users to quickly generate an immersive world on their mobile phones through a single image that not only allows for free walking, but also supports interactive operations such as "casting battles".

Click on the Lingguang App - Home page "Explore New World Models" to experience

The biggest change in the new version compared to the previous version where we could only "take a walk and have a look" is to make the AI generated world truly "playable". Users can not only stroll and explore the scene, but also perform skill operations such as attacking, archery, casting spells, and shooting like playing a 3D game. AI will generate new scene feedback in real time based on user actions.

For example, after generating a three-dimensional fantasy world through a picture of a snow mountain castle, users can transform into magic and freely stroll, or release specific skills and see the scene change with the skills. That is to say, AI generated content is no longer just static and spatial displays, but has become a dynamic world that can be explored, interacted with, and even fought.

To support this experience, the new version has introduced deep interaction capabilities and optimized multiple technical indicators, including support for 720P high-definition picture quality, smooth operation at over 30 frames per second, and reduced interaction response latency to hundreds of milliseconds. At the same time, by combining the full screen sliding to switch viewing angles, mobile phone users can obtain a more natural immersive experience.

In addition, the new version has improved character consistency, scene stability, and dynamic performance, making the details of characters, buildings, and environments more abundant and natural during long-term exploration.

The technical capability behind this upgrade comes from LingBot World 2.0, the latest open-source real-time interactive world model developed by Ant Group's Ant Lingbo Technology. Unlike traditional AI that generates fixed content, the technical goal of the world model is to generate a dynamic world that can simulate physical laws, continuously evolve, and interact with users for a long time. LingBot World 2.0 was born to explore this direction, not only supporting hourly real-time generation, 720p/60fps high-definition real-time output, and richer interactive actions and events, but also pioneering Agent driven scene dynamic evolution, keeping the entire world "alive" rather than "generated and ended".

In April of this year, Lingguang App launched the world model experience function, taking the lead in bringing world model technology to mobile devices and realizing the function of "generating interactive worlds with one image". This upgrade further expands the interactive capabilities of the world model, allowing more ordinary users to easily experience and explore new ways of AI generated worlds.

With the development of generative AI, AI is moving from "generating content" to "generating interactive worlds". Compared to traditional content generation, the world model emphasizes real-time changes and user interaction, and is considered one of the important directions for the next generation of AI content forms. It is expected to explore more application scenarios in digital content creation, gaming and entertainment, education and training, and other fields.

At present, the new world model experience feature of Lingguang App is gradually being opened to all users. Open the Lingguang App and click on the homepage "Explore New World Models" to experience it.