GLM-5.3-FlashX: Delivering inference speeds of 200 tokens/s

5x the price of 5.3 flash and the measured speed is only 75tk/s, whereas 5.3 flash measures 40tk/s. So it's either too slow or too expensive.

https://openrouter.ai/z-ai/glm-5.3-flash

https://openrouter.ai/z-ai/glm-5.3-flashx