Ant Lingbo Releases LingBot-VLA2.0: One Brain for 20 Robot Configurations
机器之心 · wechat · 2026-07-08
Ant Lingbo has officially released the new generation embodied foundation model LingBot-VLA2.0, focusing on the "one brain, many machines" approach. The pre-training covers 20 configurations from 17 mainstream domestic and international robot brands, expanding the range of motion from robotic arms to the head, waist, end-effectors, and mobile bases, covering full-body degrees of freedom. The model introduces future prediction capabilities and natively integrates the self-developed LingBot-Depth spatial perception module. In the Shanghai Jiao Tong University GM-100 evaluation system, the dual-arm robot platform's task progress score and success rate outperformed both NVIDIA GR00T N1.7 and π0.5; it also leads π0.5 in long-horizon mobile manipulation tasks. LingBot-VLA2.0 continues its open-source strategy, with model weights, training code, and technical reports released simultaneously, available on HuggingFace and the ModelScope community.
Related event: Ant Lingbo Open-Sources LingBot-VLA 2.0 for Multi-Robot Generalization(18 posts)→
More from Embodied
- NVIDIA releases MotionBricks: real-time humanoid motion generation at 15,000 FPS with 2ms latency — _akhaliq · 2026-10-26
- Claude Code finds root backdoor in Honda Civic head unit, then debloats it — IndraVahan · 2026-10-26
- Temple health wearable opens pre-orders, users ask: why wear it before benefits are proven — abhishekcode42 · 2026-10-26
- FLASH: GPU-native deformable sim learns cloth folding in minutes on one RTX 5090 — ChongZzZhang · 2026-10-26
- Luke Wroblewski spots it: iPhone Duo's corner radius mirrors the original iPhone — LukeW · 2026-10-26
- Humanoids shattered forecasts: half-marathon record in 2026, 100m faster than Bolt — ben_j_todd · 2026-10-26