AI development platform Hugging Face has announced a new robotics model called SmolVLA, designed to be highly efficient and accessible. Launched this week, SmolVLA is claimed to outperform much larger models in both virtual and real-world environments, despite its relatively compact size of 450 million parameters. Parameters, or weights, are the internal components of a model that dictate its behavior. Hugging Face's primary goal with SmolVLA is to democratize access to vision-language-action (VLA) models and speed up research toward more versatile robotic agents. The company emphasizes that SmolVLA is not only a lightweight and capable model but also a method for training and evaluating these technologies. Trained on community-shared datasets from LeRobot, SmolVLA can be run on a single consumer-grade GPU, such as those found in a MacBook, making it accessible to a broader audience. This release is part of Hugging Face’s growing push to create a low-cost ecosystem for robotics hardware and software. In 2022, the company introduced LeRobot, a suite of robotics-focused models, datasets, and tools. More recently, Hugging Face acquired Pollen Robotics, a French startup, and has begun offering various affordable robotics systems, including humanoids, for researchers and hobbyists. One of the unique features of SmolVLA is its support for an "asynchronous inference stack." This allows the model to handle a robot’s actions independently from its sensory inputs, such as visual and auditory data. According to Hugging Face, this separation enables robots to respond more swiftly and effectively in dynamic environments, where conditions change rapidly. SmolVLA is now available for download from Hugging Face’s platform. A user on social media platform X has already reported using the model to control a third-party robotic arm, highlighting its immediate practical applications. While Hugging Face is a significant player in the open robotics field, it is not alone. Competitors like Nvidia offer a suite of tools for open robotics, and startups such as K-Scale Labs are developing components for open-source humanoids. Other notable firms in the segment include Dyna Robotics, Physical Intelligence, which is backed by Jeff Bezos, and RLWRLD. The availability of SmolVLA underscores Hugging Face’s commitment to making advanced robotics accessible to a wider audience, potentially transforming how researchers and enthusiasts approach and develop robotics projects. By providing a powerful, yet lightweight, model, Hugging Face aims to facilitate faster and more innovative advancements in the field.