While Vision–Language–Action (VLA) models map visual inputs and language instructions directly to robot actions, they often rely on costly hardware and struggle in novel or cluttered scenes. We introduce EverydayVLA, a 6-DOF manipulator that can be assembled for $300, capable of modest payloads and workspaces. A single unified model jointly outputs discrete and continuous actions, and our adaptive-horizon ensembler monitors motion uncertainty to trigger on-the-fly replanning for safe, reliable operation. On LIBERO, EverydayVLA matches state-of-the-art success rates, and in real-world tests it outperforms prior methods by 49% in-distribution and 34.9% out-of-distribution. By combining a state-of-the-art VLA with cost-effective hardware, EverydayVLA democratizes access to a robotic foundation model, and paves the way for economical use in homes and research labs alike.
We tested our model, EverydayVLA, against OpenVLA and OpenVLA-OFT trained on our dataset.
We tested the models on objects seen in training: block, ball, and rock.
Additionally, we tested the models on other actions seen in training.
We also tested the models on objects not seen in training: a blue flask and a creeper keychain.
We tested with static and dynamic distractions
