Hugging Face's logo

FriendliAI
/
MiMo-Embodied-7B

Image-to-Text
Transformers
Safetensors
qwen2_5_vl
text-generation-inference
Model card Files
xet
Community
MiMo-Embodied-7B
File size: 3,892 Bytes 
d140ed3
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
 
---
base_model:
- XiaomiMiMo/MiMo-Embodied
library_name: transformers
license: mit
---

<div align="center">
  <img src="./assets/xfmlogo.svg" width=600>
</div>

<br/>

<div align="center" style="line-height: 1;">
  |
  <a href="https://huggingface.co/XiaomiMiMo/MiMo-Embodied-7B" target="_blank">🤗 HuggingFace</a>
  &nbsp;|
  <a href="https://arxiv.org/abs/2511.16518" target="_blank">📔 Technical Report</a>
  &nbsp;|
  <br/>
</div>

## I. Introduction

**MiMo-Embodied**, a powerful cross-embodied vision-language model that shows state-of-the-art performance in both **autonomous driving** and **embodied AI tasks**, the first open-source VLM that integrates these two critical areas, significantly enhancing understanding and reasoning in dynamic physical environments.

<div align="center">
  <img src="./assets/fig1.svg" width=800>
</div>


## II. Model Capabilities

<div align="center">
  <img src="./assets/fig2.svg" width=800>
</div>

## III. Model Details

<div align="center">
  <img src="./assets/fig3_img.png" width=800>
</div>

## IV. Evaluation Results

MiMo-Embodied demonstrates superior performance across **17 benchmarks in three key embodied AI capabilities: Task Planning, Affordance Prediction, and Spatial Understanding**, significantly surpassing existing open-source embodied VLM models and rivaling closed-source models.

Additionally, MiMo-Embodied excels in **12 autonomous driving benchmarks across three key capabilities: Environmental Perception, Status Prediction, and Driving Planning**—significantly outperforming both existing open-source and closed-source VLM models, as well as proprietary VLM models.

Moreover, evaluation on **8 general visual understanding benchmarks** confirms that MiMo-Embodied retains and even strengthens its general capabilities, showing that domain-specialized training enhances rather than diminishes overall model proficiency.

### Embodied AI Benchmarks

####  Affordance & Planning

<div align="center">
  <img src="./assets/table2.png" width=800>
</div>

#### Spatial Understanding

<div align="center">
  <img src="./assets/table3.png" width=800>
</div>


### Autonomous Driving Benchmarks

#### Single-View Image & Multi-View Video

<div align="center">
  <img src="./assets/table4.png" width=800>
</div>


#### Multi-View Image & Single-View Video

<div align="center">
  <img src="./assets/table5.png" width=800>
</div>

### General Visual Understanding Benchmarks

<div align="center">
  <img src="./assets/table8.png" width=800>
</div>

> Results marked with \* are obtained using our evaluation framework.


## V. Case Visualization

### Embodied AI

#### Affordance Prediction

<div align="center">
  <img src="./assets/afford-1.svg" width=800>
</div>

#### Task Planning

<div align="center">
  <img src="./assets/planning-1.svg" width=800>
</div>

#### Spatial Understanding

<div align="center">
  <img src="./assets/spatial-1.svg" width=800>
</div>

### Autonomous Driving

#### Environmental Perception

<div align="center">
  <img src="./assets/ad-perception-1.svg" width=800>
</div>

#### Status Prediction

<div align="center">
  <img src="./assets/ad-prediction-1.png" width=800>
</div>

#### Driving Planning

<div align="center">
  <img src="./assets/ad-planning-1.png" width=800>
</div>

### Real-world Tasks

#### Embodied Navigation

<div align="center">
  <img src="./assets/figure_navigation.svg" width=800>
</div>

#### Embodied Manipulation

<div align="center">
  <img src="./assets/figure_manipulation.svg" width=800>
</div>


## VI. Citation

```bibtex
@misc{hao2025mimoembodiedxembodiedfoundationmodel,
      title={MiMo-Embodied: X-Embodied Foundation Model Technical Report}, 
      author={Xiaomi Embodied Intelligence Team},
      year={2025},
      eprint={2511.16518},
      archivePrefix={arXiv},
      primaryClass={cs.RO},
      url={https://arxiv.org/abs/2511.16518}, 
}
```