SmolVLA: A Vision-Language-Action Model for Affordable and Efficient Robotics

TL;DR

SmolVLA is a compact, efficient vision-language-action model for robotics that reduces training and deployment costs while maintaining high performance, enabling broader accessibility and real-time responsiveness in robotic applications.

Contribution

This work introduces SmolVLA, a small and efficient VLA model trained on community data, capable of running on consumer hardware with competitive performance.

Findings

Achieves comparable performance to larger VLAs despite being 10x smaller.

Can be trained on a single GPU and deployed on consumer-grade hardware.

Improves responsiveness with asynchronous inference and chunked action generation.

Abstract

Vision-language models (VLMs) pretrained on large-scale multimodal datasets encode rich visual and linguistic knowledge, making them a strong foundation for robotics. Rather than training robotic policies from scratch, recent approaches adapt VLMs into vision-language-action (VLA) models that enable natural language-driven perception and control. However, existing VLAs are typically massive--often with billions of parameters--leading to high training costs and limited real-world deployability. Moreover, they rely on academic and industrial datasets, overlooking the growing availability of community-collected data from affordable robotic platforms. In this work, we present SmolVLA, a small, efficient, and community-driven VLA that drastically reduces both training and inference costs, while retaining competitive performance. SmolVLA is designed to be trained on a single GPU and deployed on consumer-grade GPUs or even CPUs. To further improve responsiveness, we introduce an asynchronous inference stack decoupling perception and action prediction from action execution, allowing higher control rates with chunked action generation. Despite its compact size, SmolVLA achieves performance comparable to VLAs that are 10x larger. We evaluate SmolVLA on a range of both simulated as well as real-world robotic benchmarks and release all code, pretrained models, and training data.