GigaBrain-0: A World Model-Powered Vision-Language-Action Model

TL;DR

GigaBrain-0 is a versatile vision-language-action model for robots that leverages world model-generated data to enhance generalization, reduce real-world data needs, and improve robustness in complex manipulation tasks.

Contribution

The paper introduces GigaBrain-0, a novel foundation model that uses synthetic data from world models to improve robot task generalization and robustness, reducing dependence on real data.

Findings

Outperforms existing models in generalization across appearance and viewpoint variations.

Achieves superior real-world performance on dexterous and long-horizon tasks.

Includes a lightweight version suitable for edge devices like NVIDIA Jetson.

Abstract

Training Vision-Language-Action (VLA) models for generalist robots typically requires large-scale real-world robot data, which is expensive and time-consuming to collect. The inefficiency of physical data collection severely limits the scalability, and generalization capacity of current VLA systems. To address this challenge, we introduce GigaBrain-0, a novel VLA foundation model empowered by world model-generated data (e.g., video generation, real2real transfer, human transfer, view transfer, sim2real transfer data). By leveraging world models to generate diverse data at scale, GigaBrain-0 significantly reduces reliance on real robot data while improving cross-task generalization. Our approach further improves policy robustness through RGBD input modeling and embodied Chain-of-Thought (CoT) supervision, enabling the model to reason about spatial geometry, object states, and long-horizon dependencies during task execution. This leads to substantial gains in real-world performance on dexterous, long-horizon, and mobile manipulation tasks. Extensive experiments demonstrate that GigaBrain-0 achieves superior generalization across variations in appearances (e.g., textures, colors), object placements, and camera viewpoints. Additionally, we present GigaBrain-0-Small, an optimized lightweight variant designed to run efficiently on devices such as the NVIDIA Jetson AGX Orin.