World models have shown great utility in improving the task performance of embodied agents. While prior work largely focuses on pixel-space world models, these approaches face practical limitations in mobile agent settings, where predicting complex visual GUI elements in future states is often difficult. In this work, we explore an alternative formulation of world modeling for mobile agents, where state transitions are described in natural language rather than predicting screenshot pixels. First, we introduce MobileWorldBench, a benchmark that evaluates the ability of vision-language models (VLMs) to function as world models for mobile agents. Second, we release MobileWorld, a large-scale dataset consisting of 1.4M samples, that improves the world modeling capabilities of VLMs. Finally, we propose a novel framework that integrates VLM world models into the planning framework of mobile agents, demonstrating that semantic world models can directly benefit mobile agents by improving task success rates.