Z.ai's GLM-5.2 open weight model impresses Silicon Valley developers
GLM-5.2, a new open-weight AI model from Chinese startup Z.ai, has quickly become a hot topic in Silicon Valley since its launch in June 2026.
Built specifically for coding and agentic workflows, it is impressing developers by solving tough programming problems and building apps that rival those made with some of the most advanced AI models from OpenAI, Anthropic, and Google.
GLM-5.2 uses 744B parameters
What sets GLM-5.2 apart? It can handle huge codebases thanks to its one-million-token context window and packs 744 billion total parameters, with 40 billion active parameters through a Mixture-of-Experts architecture.
Developers love that it is open-weight, so they can tweak and run it themselves.
Arena AI showed GLM-5.2 ranking second on the Code Arena Frontend leaderboard, earning praise from Vercel CEO Guillermo Rauch, who posted, "Genuinely impressed, almost shocked, at how good GLM-5.2 by Z.ai is at coding. This changes things," and catching the eye of investors and tech executives.