GLM-5.2: Open Weights, Near-Frontier Intelligence — Zixuan Li, Z.ai
Zixuan Li, head of Z.ai, introduces GLM-5.2 to the world. On the hardest long-horizon coding and agentic benchmarks, it lands between Claude Opus 4.7 and 4.8.
Even its non-thinking mode beats GLM-5.1 with thinking turned on, and it leads other open-weight models on the Artificial Analysis Intelligence Index. Li also explains where the GLM name comes from (a 2021 paper that put Z.ai among the earliest LLM labs), and why GLM is more than a coding model. The core of his talk is why Z.ai releases near-frontier models as open weights. It lets enterprises and governments run models on their own servers, lets companies fine-tune for law, finance and security, and lets customers co-design where the models go next. He closes with one more thing: Z Code, Z.ai's own coding harness built for GLM-5.2 that also works with other frontier models. Related links: Z.ai: https://z.ai Timestamps: 0:00 Intro from swyx 0:47 Z.ai and GLM 2:05 Where the GLM name comes from 3:10 Intelligence beyond IQ 4:25 GLM-5.2 benchmarks vs.
The text is the source's own description of its publication. The content belongs to AI Engineer.
More to read
Lumaril, which develops AI software for optimizing industrial production by connecting to existing control systems, raised a $66M Series A at a $400M valuation
Techmeme 4 h ago
DeepSeek open sources Super-Agent as a research prototype, a proactive coding agent it says can monitor group chats for coding issues and suggest solutions
Techmeme 5 h ago
Insurance software maker Guidewire agrees to acquire Hyperexponential, which makes AI-powered insurance underwriting and pricing software, for $1.7B in cash
Techmeme 5 h ago
One word for AI agents in 2027
Nate Herk 9 h ago