Intel CEO says low token costs boost rapid uptake of Chinese AI models

Intel CEO Lip‑Bu Tan said token costs of Chinese frontier AI models are lower and adoption has been very fast.

What Happened

Tan spoke at a DBS‑organised panel on Monday, noting that Chinese models such as DeepSeek, Kimi and GLM deliver performance he described as "good enough" for many tasks. He stressed his strong belief in open‑source software, calling it the best way to drive efficiency in AI development.

Outlook

Tan added that markets will decide which models serve different use cases best, and he forecast that quantum computing could become mainstream within the next three to five years. He also said software will play a key role in using the infrastructure needed for such advanced computing efficiently.

Context

Chinese AI models are gaining traction because they cost less per token, making them attractive for developers. Open‑source approaches help spread improvements quickly across the ecosystem. Quantum‑computing advances are still years away but could reshape the hardware landscape for AI workloads.