Chinese AI startup Zhipu AI said that its newly launched GLM-5.3-Flash, the first native multimodal model in its GLM-5 series, is served entirely by 100,000 domestically produced chips, underscoring China’s push to build large-scale AI infrastructure using homegrown semiconductors.

Z.ai's GLM-5.3-Flash runs natively on Chinese AI chips, hits 63.4 on DeepSWE, and offers API access at $0.15 per million input tokens.

GLM-5.3-Flash’s lower API service rates add fresh pressure on the country’s model developers as they race to lower inference costs to gain market share