Skip to content
RobothecaBeta
Log in

RynnBrain

  • AI model
  • Humanoids and robot learning

RynnBrain 1.1 is a family of open embodied vision-language foundation models, released at 2B, 9B and 122B-A10B scales, for robots to understand scenes and plan actions.

Made by Alibaba DAMO Academy.

Source on GitHub: github.com/alibaba-damo-academy/RynnBrain (opens in a new tab). Facts read at commit 5b86419 (opens in a new tab).

Licenses

Code
Apache-2.0 (opens in a new tab)

Read by Robotheca from LICENSE (opens in a new tab) (Apache-2.0) at commit 5b86419 (opens in a new tab), on 16 Oct 2026.

Commercial use
Yes

What it does

  • Decoder-only vision-language model producing text, pointing sequences, 3D perception and contact point signals from images and instructions.
  • Grounds objects with bounding boxes and predicts 3D bounding boxes from a single RGB image with camera intrinsics.
  • Predicts instruction-conditioned contact points and in-plane orientation from an image.
  • Serves as the basis for RynnBrain-VLA, which transfers its understanding into real-robot control on Unitree G1, Astribot and Tianji-Wuji.
  • Runs through Hugging Face transformers or SGLang, with an OpenAI-compatible server and an offline engine.
  • Includes cookbooks for spatial, object, grounding, area, affordance and trajectory localization, plus contact point and 3D grounding.

Written by Robotheca from the project's README (opens in a new tab) at commit 5b86419.

Getting started

Difficulty
Advanced

Maintenance

Status
Active: a commit in the last 90 days
Newest commit
10 Aug 2026 on main
Latest release
None published on GitHub

Checked on GitHub on 16 Oct 2026.

Links

Robots that can run it

No robot on Robotheca is linked to it yet. Links come from a robot's robot.yaml, from its dependency files read as text, or from Robotheca's editors.

Robotheca lists AI models as a reference: nothing is hosted or run here, and every fact links to where it was read.