LingBot-VLA 2.0
$ 0
The second generation of Robbyant’s open robot brain, trained on about 60,000 hours of data to control arms, hands and mobile robots of many kinds.
Available on backorder
Brain Score
4
Specifications and details:
| Nationality | China |
|---|---|
| Website | https://www.robbyant.com |
| Model type | Foundation Model |
| Manufacturer | Robbyant (Ant Lingbo Technology), part of Ant Group (China) |
| Release date | July 7, 2026 |
Description
LingBot-VLA 2.0 is the second generation of Robbyant’s open vision-language-action model. Ant Group’s embodied AI company built it to move robot learning from research toward real jobs. The training data grew to about 50,000 hours of robot recordings plus 10,000 hours of first-person human video. It also covers 20 robot bodies, from fixed arms to wheeled robots and humanoids. Because each robot moves differently, Robbyant maps them all into one shared action format. In Robbyant’s tests, the model beat Pi0.5 and GR00T N1.7 on real two-arm tasks.
The model can now drive arms, hands, a waist, a head and a moving base from a single brain. As a result, robots can handle longer jobs, such as stocking a fridge or cleaning a stove. It also predicts what the scene should look like next, which helps it plan ahead. Developers can download the 6B checkpoint from Hugging Face and ModelScope under Apache-2.0. Then they can fine-tune it on their own robot data with the included tools. Robbyant sees this release as a step toward its goal of one brain for every robot.
New! 2026 Humanoid
Robot Market Report
198 pages of exclusive insight from global robotics experts — uncover funding trends, technology challenges, leading manufacturers, supply chain shifts, and surveys and forecasts on future humanoid applications.
now Google DeepMind
Contact Humanoid.guide
Website: https://technology.robbyant.com/lingbot-vla-v2
