Zhipu AI Launches Faster GLM-5.3-Flash with Up to 200 Tokens/s

Zhipu AI has announced the official launch of a faster version of its GLM-5.3-Flash model, with model code glm-5.3-flashx and inference speeds reaching up to 200 tokens per second. The model is now available to all API users, while Coding Plan users can apply through a form.

The GLM series is Zhipu AI's family of large language models. The newly launched faster version is priced at 2.5 times the standard GLM-5.3-Flash on both the Coding Plan and API, distinguishing it from the existing offering.