Zhipu launches GLM-5.1 high-speed API.

Zhipu Open Platform has released GLM-5.1 high-speed API, achieving a model output speed of 400 tokens/s, setting a new global record for large-scale model API speed. Developed jointly by Zhipu and the TileRT team, GLM-5.1-highspeed is the first domestically produced large-scale model API to achieve both flagship-level capabilities and extremely low latency. It is suitable for scenarios with extremely high latency requirements, such as AI programming, real-time interaction, and real-time voice processing, and is currently available to select enterprise customers.