A world model is a learned predictive model of an environment. Given the current state or observation history and an action, it predicts the next state, observation or latent representation, and often a reward. An agent can then plan, imagine rollouts, or learn a policy inside the model instead of only in the real environment. Ha and Schmidhuber's "World Models" (2018) is the reference point for the modern term: they trained an agent "entirely inside of its own hallucinated dream generated by its world model" and transferred the policy back to the actual environment. Learning dynamics models for control is older, under the name model-based reinforcement learning.
World models differ in what they predict. TD-MPC, available in LeRobot as TD-MPC, plans "in the latent space of a learned implicit (decoder-free) world model" (Hansen et al.) and never generates images. Video world models predict future frames. Several policies in this matrix use them. LeRobot describes LingBot-VA as an autoregressive video-action world-model policy that interleaves predictions of future video latents and robot actions. FastWAM is a "World Action Model" that keeps video modeling during training but predicts actions directly at inference. VLA-JEPA adds a V-JEPA2 video world model as a training-only loss.
A common misconception is that "world model" implies photorealistic video generation, or that a policy trained with a world model runs one on the robot. Neither holds. FastWAM and VLA-JEPA drop the video-prediction path at inference, while LingBot-VA keeps it, and that difference matters for inference latency and memory on a given compute target. A world model also does not remove the sim-to-real problem; it is itself a learned simulator with its own errors.