Robbyant, an embodied AI company under Ant Group in Shanghai, said on July 8 that it had upgraded and open-sourced LingBot-VLA 2.0.

The release followed its July 6 launch of LingBot-Depth 2.0 and LingBot-Vision, expanding its robotics models for control and spatial perception.

Robbyant is Any Group’s robotics AI business, focused on robotics software and hardware.

Vision-language-action (VLA) models combine visual, language, and robot control in a single AI systems.

Robbyant said LingBot-VLA 2.0 is undergoing commercial pilot testing in retail sorting, logistics, and industrial automation.

It said LingBot-Depth 2.0 was trained on 150 million samples and ranked first in 12 of 16 depth-completion benchmarks. Meanwhile a LingBot-Depth technical paper evaluated the model on public datasets including NYUv2, iBims, DIODE, and ETH-SfM.

LingBot-Vision was released in four versions, and Robbyant said its model weights were open-sourced.

🔗 Source: Ant Group