It's good that they exist, and I hope they catch up, but if you don't have origin constraints for your use case I don't see why you would chose their models today.
[0]: On the only benchmark they both published performance results - SWE-bench Verified -they are within a margin of error Mistral 77.6 vs Qwen 77.2.
What I am trying to get at is that the frontier is great, but you can be fine with less as well.
Not for the EU. Given the political importance of LLMs and the talent pool in France (let alone rest of the EU), I fully expect them to catch up.