Vision-Language-Action (VLA) models have shown promising capabilities for embodied intelligence, but most existing approaches rely on text-based chain-of-thought reasoning where visual inputs are treated as static context. This limits the ability of the model to actively revisit the environment and resolve ambiguities during long-horizon tasks. We propose VLA-Thinker, a thinking-with-image reasoning framework that models perception as a dynamically invocable reasoning action. To train such a system, we introduce a two-stage training pipeline consisting of (1) an SFT cold-start phase with curated visual Chain-of-Thought data to activate structured reasoning and tool-use behaviors, and (2) GRPO-based reinforcement learning to align complete reasoning-action trajectories with task-level success. Extensive experiments on LIBERO and RoboTwin 2.0 benchmarks demonstrate that VLA-Thinker significantly improves manipulation performance, achieving 97.5% success rate on LIBERO and strong gains across long-horizon robotic tasks. Project and Codes: https://cywang735.github.io/VLA-Thinker/ .
@article{arxiv.2603.14523,
title = {VLA-Thinker: Boosting Vision-Language-Action Models through Thinking-with-Image Reasoning},
author = {Chaoyang Wang and Wenrui Bao and Sicheng Gao and Bingxin Xu and Yu Tian and Yogesh S. Rawat and Yunhao Ge and Yuzhang Shang},
journal= {arXiv preprint arXiv:2603.14523},
year = {2026}
}
Comments
We introduce VLA-Thinker, the first VLA model capable of thinking-with-image reasoning, which models visual perception as a dynamically invocable reasoning action, enabling Multimodal Embodied Chain-of-Thought