Mid‑2026 data show U.S. and global companies increasingly routing workloads to Chinese AI models to cut costs. Providers such as DeepSeek and Z.ai are cited as offering performance comparable to leading U.S. systems at significantly lower prices, and Chinese options have taken a much larger token share on intermediaries like OpenRouter since the start of the year.
Historically, lower‑cost technology platforms that deliver acceptable performance have often gained share over time, as seen with public cloud versus on‑prem IT, ARM‑based mobile chips versus x86 in smartphones, and low‑fee index funds versus high‑fee active mutual funds. When this pattern holds, spending migrates toward cheaper providers, pressuring higher‑cost incumbents on pricing and margins.
In the current AI cycle, Chinese platform players such as Baidu (BIDU), Alibaba (BABA), Tencent (0700.HK) (TCEHY), and Huawei‑linked ecosystems are positioned to compete on price in models and cloud infrastructure. Any resulting shift in revenue and workload share across global AI and cloud providers, however, remains conditional on factors such as sustained cost gaps, customer switching costs, and especially geopolitics and regulation, which are unusually prominent constraints in this cross‑border AI market.
References
- https://www.cio.com/article/4194040/chinese-ai-models-are-gaining-ground-in-the-u-s-attracting-companies-with-lower-prices.html
- https://www.cnbc.com/2026/07/07/chinese-ai-models-costs-us-openai-anthropic.html
- https://techstrong.ai/articles/u-s-tech-companies-turn-to-chinese-ai-models-as-domestic-costs-skyrocket/
- https://www.arm.com/company/news/2020/02/arm-190-billion-arm-based-chips-shipped