zai-org/GLM-5

GLM-5 is a large language model developed by zai-org, featuring 744 billion parameters (40B active) and trained on 28.5 trillion tokens. It integrates DeepSeek Sparse Attention to reduce deployment costs while maintaining long-context capacity. This model is specifically designed for complex systems engineering and long-horizon agentic tasks, demonstrating best-in-class performance among open-source models in reasoning, coding, and agentic benchmarks.