Z.ai releases open-weight GLM-5.3, beating GPT-5.

GLM-5.3 hit 84.5% on the CyberGym vulnerability benchmark and scored 60 on the Artificial Analysis Intelligence Index, tying for top among open-weights models. It outperforms GPT-5.6 and Claude Fable 5 on agentic benchmarks, with a low hallucination rate. Available on Databricks, Together AI, Modular Cloud, and Perplexity Computer.