TD-MPC2: Scalable, Robust World Models for Continuous Control

Nicklas Hansen, Hao Su, Xiaolong Wang
2/12/2026

No official code found. Best implementation: nicklashansen/tdmpc2 (905 stars, MIT, last commit 2025-05-21).

Abstract

TD-MPC is a model-based reinforcement learning (RL) algorithm that performs local trajectory optimization in the latent space of a learned implicit (decoder-free) world model. In this work, we present TD-MPC2: a series of improvements upon the TD-MPC algorithm. We demonstrate that TD-MPC2 improves significantly over baselines across 104 online RL tasks spanning 4 diverse task domains, achieving consistently strong results with a single set of hyperparameters. We further show that agent capabilities increase with model and data size, and successfully train a single 317M parameter agent to perform 80 tasks across multiple task domains, embodiments, and action spaces. We conclude with an account of lessons, opportunities, and risks associated with large TD-MPC2 agents. Explore videos, models, data, code, and more at https://tdmpc2.com

View on arXivView PDF

Code Implementations(1)

nicklashansen/tdmpc2Community82%
MIT

Code for "TD-MPC2: Scalable, Robust World Models for Continuous Control"

905200Shell, PythonOct 26, 20231 years agoMIT
reinforcement-learningroboticsworld-model

Similar Papers

TD-MPC-Opt: Distilling Model-Based Multi-Task Reinforcement Learning Agents

85% match1 repo22025

Discrete Codebook World Models for Continuous Control

85% match1 repo332025

PWM: Policy Learning with Multi-Task World Models

85% match2024

Learning Massively Multitask World Models for Continuous Control

85% match2025

TD-M(PC)$^2$: Improving Temporal Difference MPC Through Policy Constraint

84% match1 repo892025

Accelerating Model-Based Reinforcement Learning with State-Space World Models

83% match2025

Cite this paper

@article{hansen2026tdmpc,
  title  = {TD-MPC2: Scalable, Robust World Models for Continuous Control},
  author = {Nicklas Hansen and Hao Su and Xiaolong Wang},
  year   = {2026},
  eprint = {2310.16828},
  archivePrefix = {arXiv},
  url    = {https://arxiv.org/abs/2310.16828},
  journal = {ICLR 2023 2023}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2310.16828.svg)](https://codeofpaper.com/paper/2310.16828)

Discussion