Global AI computing power surges as token usage posts first reversal in two years

Global token consumption for AI large models has fluctuated recently, while the market landscape has undergone significant shifts.
According to the latest industry data, after ten consecutive weeks of growth, total token usage for global AI large models declined for two weeks in a row. For the week of April 13–19, usage held at 20.6 trillion tokens. Notably, weekly token usage for Chinese AI large models — which had been strong — dropped 23.77% week-over-week to 4.44 trillion tokens.
In contrast, token usage in the U.S. AI large model market rose 20.62% over the same period to 4.91 trillion tokens, marking the first time in two months that U.S.-based models have outpaced their Chinese counterparts in activity levels.
Industry experts attribute this shift — a decline in Chinese token usage and a rise in U.S. token usage — to the global surge in computing power costs. Since March, major Chinese cloud providers such as Alibaba Cloud, Tencent, and Baidu have raised prices for large model services one after another. Meanwhile, domestic AI unicorn Zhipu has completed its third price adjustment this year. Overseas, AI giant Anthropic has also updated its model pricing.
Although token usage has temporarily dipped due to price fluctuations and market adjustments, the industry remains optimistic. Most analysts believe that as AI applications continue to deepen, global token consumption will keep growing over the long term, and the current fluctuations are seen as a natural market correction amid the reshaping of computing power costs.
Related article
Former OpenAI Researcher Barrett Zoph Returns to Google to Support Gemini Development
Barret Zoph, who previously abruptly left and joined OpenAI due to a disagreement with Thinking Machines Lab CEO Miriam Mularati, has once again changed employers. Alphabet's Google announced that Zoph will return as Vice President of Research, bring
How Recursive Intelligence Raised $335M at $4B Valuation in 4 Months
The co-founders of startup Ricursive Intelligence seemed destined to be co-founders.Anna Goldie, CEO, and Azalia Mirhoseini, CTO, are so well-known in the AI community that they were among those AI engineers who “got those weird emails from Zuckerber
AI race won’t wait, nor will this TechCrunch Disrupt 25% discount
Every founder building in AI right now is asking the same question: Are we moving fast enough, and effectively enough? TechCrunch Disrupt 2026 puts you in the same room as the people wrestling with that existential question at the highest levels. Fro
Related Special Topic Recommendations
Comments (0)
0/500

Global token consumption for AI large models has fluctuated recently, while the market landscape has undergone significant shifts.
According to the latest industry data, after ten consecutive weeks of growth, total token usage for global AI large models declined for two weeks in a row. For the week of April 13–19, usage held at 20.6 trillion tokens. Notably, weekly token usage for Chinese AI large models — which had been strong — dropped 23.77% week-over-week to 4.44 trillion tokens.
In contrast, token usage in the U.S. AI large model market rose 20.62% over the same period to 4.91 trillion tokens, marking the first time in two months that U.S.-based models have outpaced their Chinese counterparts in activity levels.
Industry experts attribute this shift — a decline in Chinese token usage and a rise in U.S. token usage — to the global surge in computing power costs. Since March, major Chinese cloud providers such as Alibaba Cloud, Tencent, and Baidu have raised prices for large model services one after another. Meanwhile, domestic AI unicorn Zhipu has completed its third price adjustment this year. Overseas, AI giant Anthropic has also updated its model pricing.
Although token usage has temporarily dipped due to price fluctuations and market adjustments, the industry remains optimistic. Most analysts believe that as AI applications continue to deepen, global token consumption will keep growing over the long term, and the current fluctuations are seen as a natural market correction amid the reshaping of computing power costs.
Former OpenAI Researcher Barrett Zoph Returns to Google to Support Gemini Development
Barret Zoph, who previously abruptly left and joined OpenAI due to a disagreement with Thinking Machines Lab CEO Miriam Mularati, has once again changed employers. Alphabet's Google announced that Zoph will return as Vice President of Research, bring
How Recursive Intelligence Raised $335M at $4B Valuation in 4 Months
The co-founders of startup Ricursive Intelligence seemed destined to be co-founders.Anna Goldie, CEO, and Azalia Mirhoseini, CTO, are so well-known in the AI community that they were among those AI engineers who “got those weird emails from Zuckerber
AI race won’t wait, nor will this TechCrunch Disrupt 25% discount
Every founder building in AI right now is asking the same question: Are we moving fast enough, and effectively enough? TechCrunch Disrupt 2026 puts you in the same room as the people wrestling with that existential question at the highest levels. Fro





Home






