Chinese AI Models Take Nearly 90% of Global Open-Source Model Spending

A recent report from Silicon Valley venture firm Menlo Ventures shows that Chinese AI models keep gaining ground in the enterprise open-source market.

The data shows Chinese AI models, led by Moonshot AI's Kimi series, now account for 89% of global enterprise spending on open-source models. Kimi ranks first with a 61% share, DeepSeek follows at 15%, and Zhipu's GLM takes third place with 11%.

By comparison, Meta's Llama series is the top Western model in this field, but it holds only a 7% share of enterprise open-source model spending, a clear gap with Chinese models.