European businesses are increasingly adopting open-weight Chinese AI models driven by cost, competitive performance, and the desire for greater operational control — and analysts argue that locally hosting these systems can strengthen Europe's technological sovereignty by reducing dependence on U.S. hyperscalers, even as it risks creating new dependencies on Chinese model ecosystems.
What is driving European adoption
- Radical cost advantage. Chinese open-weight models run 60% to 90% cheaper than flagship models from OpenAI and Anthropic. DeepSeek-V4-Pro, for example, is priced roughly 12× below GPT-5.5 at comparable benchmark intelligence
![]()
. This cost gap is the primary wedge: "price is doing the work here," as one infrastructure executive put it
.
- Competitive, sometimes superior performance. DeepSeek V3.2-Speciale scores 96.0% on AIME 2025, outperforming GPT-5 High; GLM-5.2 outperforms GPT-5.5 on SWE-bench Pro at one-sixth the cost
. By May 2026, Chinese open-weight models accounted for roughly 61% of all tokens consumed on OpenRouter, the largest neutral LLM router
.
- Permissive licensing and availability. Models ship under MIT, Apache 2.0, or Apache-style commercial licenses, allowing enterprises to download, fine-tune, and self-host without API gatekeepers . Qwen surpassed 1 billion cumulative downloads on Hugging Face by March 2026 .