New report shows chinese AI models dominate global enterprise open-source spending, led by Kimi

A recent report released by venture capital firm Menlo Ventures indicates that Chinese artificial intelligence models account for an overwhelming 89% of global enterprise spending on open-source models. The analysis highlights the dominance of domestic players. Moonshot AI's Kimi series leads the spending with a 61% share, followed by DeepSeek at 15%, and GLM from Zhipu AI securing the third spot with 11% of the expenditure. In contrast, the Llama series from Meta Platforms, while being the most dominant Western model in this sector, captured only 7% of the total enterprise spending on open-source models.