A surge of open-weight AI models from Chinese labs is narrowing the gap with leading closed systems like Claude, ChatGPT, and Gemini, especially in coding benchmarks. Z.ai’s GLM‑5.1 and Moonshot’s ...