Reinforcement learning (RL) for large-scale Vision-Language-Action (VLA) models faces significant challenges in computational efficiency and data acquisition. We propose AcceRL, a fully asynchronous and decoupled RL framework designed to eliminate synchronization barriers by physically isolating training, inference, and rollouts. Crucially, AcceRL is the first to integrate a plug-and-play, trainable world model into a distributed asynchronous RL pipeline to generate virtual experiences. Experiments on the LIBERO~\cite{liu2023libero} benchmark demonstrate that AcceRL achieves state-of-the-art (SOTA) performance. Systematically, it exhibits super-linear scaling in throughput and highly efficient hardware utilization. Algorithmically, the world-model-augmented variant delivers unprecedented sample efficiency and robust training stability in complex control tasks. Code is publicly available at https://github.com/distanceLu/AcceRL.
@article{arxiv.2603.18464,
title = {AcceRL: A Distributed Asynchronous Reinforcement Learning and World Model Framework for Vision-Language-Action Models},
author = {Chengxuan Lu and Shukuan Wang and Yanjie Li and Wei Liu and Shiji Jin and Fuyuan Qian and Peiming Li and Baigui Sun and Yang Liu},
journal= {arXiv preprint arXiv:2603.18464},
year = {2026}
}