LET-KUAVO-VLA-1.0

LET-KUAVO-VLA-1.0 is an open vision-language-action (VLA) model released by Leju Robotics on Hugging Face for the KUAVO humanoid robot platform. The model repository (LejuRobotics/LET-KUAVO-VLA-1.0-models, created August 24, 2026) provides safetensors model weights together with a video preprocessor configuration; access is gated. A companion dataset, LET-KUAVO-VLA-1.0-Dataset (created August 23, 2026, Apache-2.0 license, gated), contains 5,183 video files. It is a video-modality dataset associated with the KUAVO VLA release. Leju Robotics is the developer of the KUAVO humanoid robot series, and the release is part of the company's open-scale humanoid robot platform strategy.