On August 10, according to LatePost, the registered user base of Zhipu's MaaS open platform has approached 7 million, an increase of about 2 million since early July, with enterprise customers reaching 23,000. Its developer product ZCode surpassed 1 million users within a month of launch. The report said Zhipu's ARR (Annual Recurring Revenue) has grown about 15 times this year. Market sources revealed that Zhipu's current ARR may reach $2 billion, but this figure has not been officially confirmed. As model call demand grows rapidly, Zhipu is expanding its domestic computing infrastructure. It is understood that Zhipu has activated more than 50,000 domestic AI computing chips to alleviate growing inference demand. Earlier, there were market rumors that Zhipu had built a 1GW domestic AI computing infrastructure. Zhipu's growth is mainly driven by the explosive demand in programming scenarios. After the release of the GLM-5 model, the company's ARR grew rapidly, and the API business has become the core revenue source. Zhipu also continues to optimize inference efficiency, improving computing utilization through technologies such as KV cache splitting and inference network architecture. In addition, Zhipu launched a high-speed API in July, generating 400 tokens per second, about 8 times faster than the regular API. On July 31, the company reopened the purchase of Coding Plan and significantly adjusted prices, with the Lite version's monthly fee rising to 118 yuan. Industry insiders believe that as AI programming demand continues to grow, inference efficiency optimization will become the core competitiveness of large model companies. Whoever can generate more tokens under the same computing power will gain advantages in cost and commercialization. At present, Chinese large model companies such as Zhipu, Kimi, and DeepSeek are all facing computing bottlenecks and model competition pressure. The market expects multiple companies to release new-generation models in August, and competition in AI large models will further intensify.