On September 18, 2026, Zhipu proudly announced the launch of its latest innovation, the GLM-5.3-FlashX. This cutting-edge model boasts an impressive inference speed, reaching up to 200 tokens per second—marking a substantial fivefold increase compared to its forerunner. Alongside this performance boost, the pricing structure has been adjusted, now set at 2.5 times the previous rate. Moreover, developers and businesses can now conveniently access this powerful tool through the newly available API.
