MolmoAct2 is an open vision-language-action model for robot control. It builds on Molmo2-ER and attaches a flow-matching continuous action expert that conditions on the VLM key-value cache through a per-layer connection.

This checkpoint is fine-tuned on the SO-100/101 mixture with absolute joint-pose control and annotated language instructions. It is intended for both further fine-tuning and SO-100/101 policy inference.

Use this checkpoint for SO-100/101 inference or for further fine-tuning. Dataset normalization metadata is stored in norm_stats.json. pass norm_tag="so100_so101_molmoact2" at inference time.

Continuous action prediction is the intended and recommended inference mode. Discrete action prediction is exposed for parity and debugging, but we use continuous actions by default.

This sample comes from Beegbrain/pick_lemon_and_drop_in_bowl, episode 0, frame 0. Camera order for this checkpoint does not matter. random camera order is acceptable.

MolmoAct2 was trained with mixed precision. For our reported experiments, we ran inference in float32. This path uses the most GPU memory: roughly 26GB with CUDA graph enabled, or around 24GB without CUDA graph.

If you have a GPU with less memory, you can run inference with bfloat16 instead:

Using bfloat16 is much more memory efficient and can run under 16GB of GPU memory in our tests. It usually does not hurt performance much.

Images may be PIL images or RGB arrays. Camera order does not need to be fixed for this checkpoint. random camera order is acceptable. state is the raw robot state, and actions are returned in robot scale.

normalize_language=True is the default. It lowercases the task string and removes trailing sentence punctuation to match training preprocessing. Set it to False if you need to preserve the task text exactly.

enable_cuda_graph=True is the default. The first few calls can be slow because the model warms up and captures CUDA graphs. run several random warm-up calls before measuring deployment latency. num_steps controls the continuous flow solver and defaults to the checkpoint config value, 10.

Depth reasoning is disabled for this checkpoint. Calling enable_depth_reasoning=True will raise an error.

Discrete action inference requires a caller-provided action tokenizer. It is not saved in this repository. Discrete mode decodes action tokens directly. the continuous action expert is not used.

MolmoAct2 generate robot actions from visual observations and language instructions, but their behavior may vary across embodiments, environments, and hardware configurations. Users should carefully validate model outputs before deployment, especially when operating physical robots or other actuated systems. Where possible, actions should be monitored through interpretable intermediate outputs (adaptive depth map), simulation rollouts, action limits, or other safety checks before execution on hardware. The model’s action space should be bounded by the training data, robot controller limits, and task-specific safety constraints, including limits on speed, workspace, torque, and contact force. Users should follow the hardware manufacturer’s safety guidelines, use appropriate emergency-stop mechanisms, and operate the system only in a safely configured environment with human supervision.