According to CoinMeta, Zhipu (02513. hk) announced that it has released the GLM-5.3 model today. This model uses the same base model as GLM-5.2, and all performance improvements come from post-training optimizations. Compared to GLM-5.2, GLM-5.3 performs better in complex coding and long-term tasks, boasting a more powerful coding capability, making it the most functional open-source weight model currently available. In internal Z.ai code benchmark tests, its performance has improved by 50% compared to GLM-5.2. Moreover, during large-scale deployment after training, the development speed of its network capabilities exceeded expectations. On the Cybergym platform, GLM-5.3 leads in vulnerability detection, with the most significant improvements in leveraging the later stages of the chain. In utilization benchmark tests, its performance is more than twice that of GLM-5.2. Once the security assessment and reinforcement work are completed, Zhipu will announce the weights two weeks after the release.