glm-5.3-flashx

GLM-5.3-FlashX is the high-speed inference variant of GLM-5.3-Flash, preserving native multimodal understanding and a one-million-token context window while delivering generation speeds of up to 200 tokens per second through serving and inference optimizations. It is designed for latency-sensitive workloads such as interactive coding, tool-using agents, visual understanding, and document workflows.