Key TakeawaysChinese AI large models recorded 4.69 trillion tokens in weekly call volume for the week of March 9–15, up 11.83% week-on-weekUS AI large models saw call volume decline 9.33% to 3.294 trillion tokens in the same periodChina has now outpaced the US in weekly AI model call volume for two consecutive weeksThe top three models globally by call volume last week were all Chinese AI large modelsMysterious new model Hunter Alpha debuted at seventh place with 0.666 trillion tokens, launched March 11 with a 1 million token context window and trillions of parametersChina's AI large models have surpassed their US counterparts in weekly usage volume for a second straight week, according to the latest data from AI model aggregator OpenRouter, covering the period from March 9 to March 15.Chinese models collectively processed 4.69 trillion tokens last week, an 11.83% increase from the prior week. US models, by contrast, recorded 3.294 trillion tokens, a sequential decline of 9.33%. The top three models globally by call volume were all Chinese, marking a notable shift in the competitive landscape of large language model deployment.Hunter Alpha Debuts With No Clear OriginAmong last week's notable entrants is Hunter Alpha, a model that ranked seventh globally with 0.666 trillion tokens in weekly call volume despite having launched on March 11 -- just days before the reporting period closed.According to its OpenRouter listing, Hunter Alpha is built specifically for agent applications, features a context length of one million tokens, and is described as an intelligent model with trillions of parameters. Its developer identity and backing remain undisclosed, adding to its profile as one of the more closely watched new entrants in the current AI model race.The back-to-back weeks of Chinese dominance in raw call volume come as US AI developers including OpenAI, Anthropic, and Google continue to compete for enterprise and developer mindshare, while Chinese players have rapidly expanded both model capability and user adoption.