Ant Group's Lingbo Technology open-sourced the real-robot post-training toolchain for its embodied AI model, LingBot-VLA. The release aims to simplify deployment on diverse robots by providing tools for data processing, training, evaluation, and accelerated deployment. Pre-trained on 20,000 hours of real-robot data, LingBot-VLA reportedly outperforms benchmarks and enables task adaptation with minimal demonstration data. The model and code are available on GitHub, Hugging Face, and ModelScope.