---
title: Xiaomi-Robotics-0-Calvin-ABCD_D
canonical_url: "https://www.modelscope.ai/models/XiaomiRobotics/Xiaomi-Robotics-0-Calvin-ABCD_D"
md_url: "https://www.modelscope.ai/models/XiaomiRobotics/Xiaomi-Robotics-0-Calvin-ABCD_D.md"
repository: XiaomiRobotics/Xiaomi-Robotics-0-Calvin-ABCD_D
last_updated: 2026-07-15
license: apache-2.0
model_type:
  - mibot
architectures:
  - MiBoTForActionGeneration
parameters: 4.7B
tensor_type:
  - BF16
library_name:
  - pytorch
  - transformer
  - safetensors
frameworks:
  - pytorch
downloads: 127
stars: 0
tags:
  - robotics
  - vision-language-action
  - vla
---

# Xiaomi-Robotics-0-Calvin-ABCD_D

> Xiaomi-Robotics-0-Calvin-ABCD_D - An open-source model by XiaomiRobotics on ModelScope. Xiaomi-Robotics-0 is a state-of-the-art vision-language-action (VLA) model with 4.7B parameters, specifically engineered for high-performance robotic reasoning and…

XiaomiRobotics/Xiaomi-Robotics-0-Calvin-ABCD_D is a 4.7B-parameter machine learning model on ModelScope. licensed under apache-2.0.

- **Repository**: XiaomiRobotics/Xiaomi-Robotics-0-Calvin-ABCD_D
- **License**: apache-2.0
- **Parameters**: 4.7B
- **Tags**: robotics, vision-language-action, vla
- **Downloads**: 127
- **Stars**: 0
- **Last updated**: 2026-07-15

Source: https://www.modelscope.ai/models/XiaomiRobotics/Xiaomi-Robotics-0-Calvin-ABCD_D

---

# Xiaomi-Robotics-0

Xiaomi-Robotics-0 is a state-of-the-art vision-language-action (VLA) model with 4.7B parameters, specifically engineered for high-performance robotic reasoning and seamless real-time execution. It is pre-trained on large-scale cross-embodiment robot trajectories and vision-language data, enabling broad action-generation capabilities across different robotic platforms.

- **Paper:** [Xiaomi-Robotics-0: An Open-Sourced Vision-Language-Action Model with Real-Time Execution](https://huggingface.co/papers/2602.12684)
- **Project Page:** [https://xiaomi-robotics-0.github.io/](https://xiaomi-robotics-0.github.io/)
- **Repository:** [https://github.com/XiaomiRobotics/Xiaomi-Robotics-0](https://github.com/XiaomiRobotics/Xiaomi-Robotics-0)

## Performance

The model achieves state-of-the-art performance across several simulation benchmarks:
- **LIBERO**: 98.7% average success rate.
- **CALVIN**: 4.80 average sequence length (ABCD→D split).
- **SimplerEnv**: 85.5% success rate on Google Robot (VM) and 79.2% on WidowX (Bridge).

## Sample Usage

Xiaomi-Robotics-0 is compatible with the Hugging Face `transformers` ecosystem. Below is an example of how to load the model and generate actions for a robotic task.

```python
import torch
from transformers import AutoModel, AutoProcessor

# 1. Load model and processor 
model_path = "XiaomiRobotics/Xiaomi-Robotics-0"
model = AutoModel.from_pretrained(
    model_path, 
    trust_remote_code=True, 
    attn_implementation="flash_attention_2", 
    dtype=torch.bfloat16
).cuda().eval()
processor = AutoProcessor.from_pretrained(model_path, trust_remote_code=True, use_fast=False)

# 2. Construct the prompt with multi-view inputs
language_instruction = "Pick up the red block."
instruction = (
    f"<|im_start|>user
The following observations are captured from multiple views.
"
    f"# Base View
<|vision_start|><|image_pad|><|vision_end|>
"
    f"# Left-Wrist View
<|vision_start|><|image_pad|><|vision_end|>
"
    f"Generate robot actions for the task:
{language_instruction} /no_cot<|im_end|>
"
    f"<|im_start|>assistant
<cot></cot><|im_end|>
"
)

# 3. Prepare inputs
# Assuming `image_base`, `image_wrist` (PIL Images) and `proprio_state` (numpy array) are loaded
inputs = processor(
    text=[instruction],
    images=[image_base, image_wrist], # [PIL.Image, PIL.Image]
    videos=None,
    padding=True,
    return_tensors="pt",
).to(model.device)

# Add proprioceptive state and action mask
robot_type = "libero" # or "calvin", "widowx", etc.
inputs["state"] = torch.from_numpy(proprio_state).to(model.device, model.dtype).view(1, 1, -1)
inputs["action_mask"] = processor.get_action_mask(robot_type).to(model.device, model.dtype)

# 4. Generate action 
with torch.no_grad():
    outputs = model(**inputs)
    
# Decode raw outputs into actionable control commands
action_chunk = processor.decode_action(outputs.actions, robot_type=robot_type)
print(f"Generated Action Chunk Shape: {action_chunk.shape}")
```

## Citation

```bibtex
@misc{robotics2026xiaomi,
  title        = {Xiaomi-Robotics-0: An Open-Sourced Vision-Language-Action Model with Real-Time Execution},
  author       = {Xiaomi Robotics},
  howpublished={\url{https://xiaomi-robotics-0.github.io}},
  year         = {2026},
  note={Project Website}
}
```