智谱发布GLM-5.3

Odaily Planet Daily reported that Zhipu AI (HK2513) officially released GLM-5.3. According to the introduction, compared with GLM-5.2, the base model of GLM-5.3 remains unchanged, but the upper bound of the model's intelligence has been greatly improved through extreme post-training scaling. GLM-5.3 has stronger programming capabilities, with a 50% improvement over GLM-5.2 in internal self-built experience evaluations, and achieved first place among open-source models in public benchmarks including Terminal Bench 3.0 and Agents' Last Exam (CLI). Zhipu AI (HK2513) also stated that model weights will be open-sourced within two weeks after release. (Yicai)