Zhipu Launches GLM-5.3-FlashX with Up to 200 Tokens/s Inference Speed

Chinese AI company Zhipu has launched the GLM-5.3-FlashX model, which reaches inference speeds of up to 200 tokens per second, offering faster performance for enterprises and developers. The model was previously released to global developers under the name "Ox Alpha" and has seen growing usage.

Built on inference capacity from 100,000 domestically produced chips, Zhipu has further invested in infrastructure and inference optimization. The GLM-5.3-Flash series has long maintained top-tier intelligence at its size, and this speed boost creates all-around competitiveness in intelligence, price, and speed.

The GLM-5.3-FlashX API is now live, supporting a 1 million token context window with image, video, file, and text inputs. Pricing is set at 2 yuan per million input tokens and 7 yuan per million output tokens, with cache hits at 0.57 yuan per million tokens.