Entering the Physical World: Robbyant Launches LingBot-VA 2.0 for Robot Control

Generative algorithms are leaving digital sandboxes. On July 10, 2026, Robbyant (backed by Chinese giant Ant Group) announced the release of `LingBot-VA 2.0`. This is a neural network architecture built specifically for Embodied AI and the control of physical objects in the real world.

While Western labs are fighting to improve the accuracy of text responses, Asian engineers are making a qualitative leap. LingBot-VA 2.0 allows high-level cognitive LLM commands to be translated directly into robot motor skills. This means machines are beginning to "understand" the physics and geometry of space, rather than just parsing pixels. The integration of large models with robotics opens up an entirely new macroeconomic segment of automation—from smart warehouse logistics to autonomous manufacturing, where the algorithm itself makes decisions regarding physical manipulations.

Source: Robbyant / Ant Group