Chinese Models Lead Enterprise Open-Weight Spending
AI summary
Menlo Ventures' 2026 State of Enterprise AI report found that Chinese models accounted for 89% of the surveyed enterprise open-weight model spending, with Kimi leading at 61%, followed by DeepSeek at 15% and GLM at 11%. Llama, described as the leading Western model, held 7%, while 36Kr said the survey covered nearly 400 US enterprise AI leaders responsible for more than $4 billion in AI spending. Reports conflict over whether the spending denominator covers US enterprises or enterprises globally.
In-depth insight
The reported spending shares make Kimi and other Chinese models significant choices in enterprise open-weight AI procurement, although the disputed survey scope limits direct comparison.