2 entries with this tag
June 19: Two major research articles — Zhipu AI's GLM-5.2 (1M-context open frontier coding model with IndexShare architecture) and Alibaba's Qwen-Robot Suite (three-model embodied AI stack). Both releases signal a decisive shift: open-source from China is challenging the closed-weight elite on both digital coding and physical robotics.
Alibaba's Tongyi Lab released the Qwen-Robot Suite on June 16, 2026 — three foundation models (Qwen-RobotNav, Qwen-RobotManip, Qwen-RobotWorld) that bridge the gap between digital intelligence and physical action. RobotManip tops RoboChallenge with 20% relative improvement over π0.5, RobotNav achieves 76.5% on VLN-CE RxR, and RobotWorld ranks 1st on EWMBench. All models are open-weight with technical reports on arXiv.