GoVLA
GOVLA
GoVLA is an embodied AI foundation model developed by AI2Robotics (智平方), a portfolio company of Yunqi Capital. First released in April 2025 as what the fund described as the world's first "full-domain, full-body" VLA (Vision-Language-Action) large model, it has since iterated through GOVLA 0.0 (RoboMamba), GOVLA 0.5 (FiS-VLA), and GOVLA 1.0 . The 0.0 version was accepted at NeurIPS 2024; the 0.5 version introduced a dual-system architecture with "heterogeneous input + asynchronous frequency" to improve control frequency and task generalization . The model powers AI2Robotics's AlphaBot series of wheeled dual-arm robots, which have entered industrial manufacturing settings including automotive, biotech, and display panel lines, with a three-year, 1,000-unit order from panel maker HKC . The company has raised over one billion RMB to advance GOVLA's capabilities and scale production, with its self-built line currently capable of roughly one thousand units annually .
AI-generated — may contain errors, please verify.
Coverage
Yunqi Capital | Early Portfolio Company "Zhi Square" Raises Over 1 Billion RMB to Accelerate Embodied AI Models and Mass Production
Capital × Industry Collaboration Accelerates Embodied Intelligence Mass Production
Yunqi Capital | "Intelligent Square" Closes New Nine-Figure RMB Funding Round, Doubling Down on Model Iteration and Production Line Expansion
Embodied Intelligence, Continuously Evolving

