Compare

DEEP Robotics · robot AI model

DR-VLA-2.0

China · 2026

Sources retrieved: 27 October 2026 (A, B, C)

3 of 7 key figures stated, as stated by the developer

Model type
World action modelA
What it controls
Not stated
Inputs
Camera Language Robot stateA
Access
Not stated
Parameters
Not stated
Base model
Qwen3-VL-2B (understanding expert)A
Licence type
Not stated

Visit website

Illustrative image — not this model

Specifications

4/14 specifications disclosed 10 not stated

Not statedThe developer publishes nothing on it

Select a source marker for source details and notes.

Release1/1 disclosed

Released13 October 2026

Input and output2/4 disclosed

Model typeWorld action model
InputsCamera Language Robot state
What it controlsNot stated
Degrees of freedom controlledNot stated

Size and speed0/2 disclosed

ParametersNot stated
Control rateNot stated

Where it runs and what it is built on1/2 disclosed

On robot or cloudNot stated
Base modelQwen3-VL-2B (understanding expert)

Openness0/3 disclosed

AccessNot stated
Licence documentNot stated
Licence typeNot stated

The developer’s own claims0/2 disclosed

Learns new tasksNot stated
Works across robotsNot stated

Notes

DEEP Robotics lists DR-VLA-2.0 as an embodied foundation model in its press release of 24 October 2026; its study page and its arXiv preprint of 13 October 2026 call the described model DR-VLA.

The same press release places Atomic Motion Coordinate (AMC), a manipulation framework with its own project page, under the DR-VLA-2.0 heading.

The robot data of DR-VLA are teleoperated navigation of a wheeled-legged robot from real-world delivery settings.

Sources

One line per source, not per figure: a source has one retrieval date.

  1. Aarxiv.org/html/2610.18373Retrieved 27 October 2026
  2. Bdeveloper.deeprobotics.cn/research/articles/dr-vla/index.htmlMain source. Retrieved 27 October 2026
  3. Carxiv.org/abs/2610.18373Retrieved 27 October 2026

Report an error or send an updated datasheet