MiMo-Embodied: X-Embodied Foundation Model Technical Report

TL;DR

MiMo-Embodied is a pioneering cross-embodied foundation model that achieves state-of-the-art results in both Autonomous Driving and Embodied AI tasks through multi-stage learning and fine-tuning.

Contribution

It introduces the first cross-embodied foundation model that integrates autonomous driving and embodied AI, setting new performance benchmarks and enabling positive transfer between domains.

Findings

Sets new records across 17 embodied AI benchmarks.

Excels in 12 autonomous driving benchmarks.

Significantly outperforms existing baselines.

Abstract

We open-source MiMo-Embodied, the first cross-embodied foundation model to successfully integrate and achieve state-of-the-art performance in both Autonomous Driving and Embodied AI. MiMo-Embodied sets new records across 17 embodied AI benchmarks in Task Planning, Affordance Prediction and Spatial Understanding, while also excelling in 12 autonomous driving benchmarks across Environmental Perception, Status Prediction, and Driving Planning. Across these tasks, MiMo-Embodied significantly outperforms existing open-source, closed-source, and specialized baselines. Our results indicate that through multi-stage learning, curated data construction, and CoT/RL fine-tuning, these two domains exhibit strong positive transfer and mutually reinforce one another. We provide a detailed analysis of our model design and training methodologies to facilitate further research. Code and models are available at https://github.com/XiaomiMiMo/MiMo-Embodied.