AMD

The first Ant Lingbo embodied AI large model challenge opened at the Bund Summit on September 14, co-organized with ModelScope and Alibaba Cloud Tianchi. It uses LingBot-VLA 2.0, a pre-trained model fully open-sourced in July 2026, as its technical base and is open to enterprises, universities, and individual developers worldwide.

The online preliminary round runs through October 26, with a required RoboTwin 2.0 simulation task, an optional real-robot innovation task, and dedicated cloud compute support. Finalists will join an offline real-robot hackathon in Shanghai from November 13 to 15.

Ant Lingbo has open-sourced a 1.3B-parameter lightweight LingBot-World 2.0 world model, following the 14B version released in July. It runs real-time world generation on a single consumer GPU. Unlike a pruned checkpoint, it was trained end-to-end with causal pretraining using MoBA attention masks, consistency distillation, and DMD distribution matching distillation.

The release includes the causal pretraining base and a bidirectional teacher for community post-training, distillation, and scene adaptation. AMD senior VP Jack Huynh demonstrated LingBot-World 2.0 during his IFA Berlin opening keynote.