MolmoAct
$ 0
Multimodal VLA from the Molmo family adapted for robot action prediction; open weights with strong zero-shot generalisation
Available on backorder
Brain Score
4
Specifications and details:
| Nationality | US |
|---|---|
| Website | https://allenai.org |
| Model type | Foundation Model |
| Manufacturer | Allen Institute for AI (Ai2) |
| Release date | 2025 |
Description
MolmoAct is a foundation model that combines vision, language, and action into a single unified system. Instead of relying on predefined routines, it interprets visual and contextual information to plan movements step by step. Consequently, robots adapt quickly to new situations and unfamiliar environments. Moreover, its strong zero-shot generalization allows it to perform tasks it has not explicitly trained for, making it applicable across a wide range of real-world scenarios.
In addition, MolmoAct emphasizes real-time responsiveness and smooth coordination. Therefore, robots adjust actions dynamically as conditions change, without needing complex reprogramming. Its multimodal design enables understanding of both context and intent, improving interaction with objects and environments. Because the model provides open weights, developers can refine and expand its capabilities over time. As a result, MolmoAct supports ongoing progress toward general-purpose and adaptable robotic intelligence.
New! 2026 Humanoid
Robot Market Report
198 pages of exclusive insight from global robotics experts — uncover funding trends, technology challenges, leading manufacturers, supply chain shifts, and surveys and forecasts on future humanoid applications.
now Google DeepMind
Contact Humanoid.guide
Website: https://allenai.org
