MolmoBot

AI model
Humanoids and robot learning

MolmoBot is a research project for training robot manipulation policies in large-scale simulation that transfer zero-shot to real robots. It includes the MolmoBot-DROID policy and related data tools.

Made by Ai2, United States (checked on github.com/allenai (opens in a new tab)).

Source on GitHub: github.com/allenai/MolmoBot (opens in a new tab). Facts read at commit 33c0ca7 (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 33c0ca7 (opens in a new tab), on 16 Oct 2026.
Weights: Apache-2.0 (opens in a new tab)
Read on huggingface.co/allenai/MolmoBot-DROID (opens in a new tab).
Commercial use: Yes

What it does

Trains manipulation policies in simulation that transfer to novel real-world scenes, objects, and camera viewpoints without retraining.
Provides MolmoBot-DROID, a policy that runs on a DROID platform using only a wrist camera and one exterior camera.
Offers an interactive simulation demo where users can modify scenes and tasks to test policy behavior.
Includes instructions for serving the policy and running it on a DROID or polymetis setup with a sample language task.
Supplies scripts to download the MolmoBot training dataset, validate and filter trajectories, and compute aggregate data statistics.

Written by Robotheca from the project's README (opens in a new tab) at commit 33c0ca7.

Getting started

Difficulty: Advanced
How to start: The project's website (opens in a new tab)

Maintenance

Status: Maintained: a commit in the last 12 months
Newest commit: 27 Apr 2026 on main
Latest release: None published on GitHub

Checked on GitHub on 16 Oct 2026.

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.