GLM-5.3-FlashX is the high-speed variant of Z.ai's GLM-5.3-Flash, a native multimodal model delivering inference speeds of up to 200 tokens/s. Built on the same hybrid sparse and linear attention architecture...
by Z-ai|1M context|$0.37/M input tokens|$1.25/M output tokens
Endpoints
Available providers for this model, with details on pricing, context limits, and real-time health metrics.