As discussed in the comments QUIC is not yet formally standardised so it is unsurprising it is not available in most tools. None of the major web servers (e.g. Apache, Nginx or IIS) and even indicated they are working on it yet. It is due to be completed in July and then submitted for standardisation which will take a few months after that. After that I would expect implementations to start becoming available.
Google invented QUIC and do have a version on their servers and in Chrome. This form the basis of the QUIC that will be standardised but the two have diverged and are not compatible. So you could implement a version of Google QUIC if you wanted to and some servers like LiteSpeed and some CDNs like Akamai do this. As do Google themselves on their Cloud Platform. They basically do this by reverse engineering the open source Google Chrome code. Also as Google iterates QUIC and stops supporting the old versions they must keep up or it will stop working. Eventually Google QUIC will be deprecated and then retired once the IETF standardised QUIC comes out.
QUIC is also incredibly complex! Implementing it will not be easy and will take considerable effort and time. It is not as simple as finding the HTTP code and copying and pasting it and changing a couple of things. It’s a massive whole new protocol that reimplements parts of TCP, TLS and HTTP/2. HTTP/3 then is what is left over from HTTP/2 and needs to be implemented as well as QUIC to be useful.
Finally the the impact of QUIC might not be as large as you think. QUIC is an evolution of HTTP/2 and fixes one edge case where HTTP/2 can be slower than HTTP/1.1 if there is very high packet loss. Apart from this scenario the initial versions of QUIC will be very similar to HTTP/2 and TLSv1.3 which are available now. One of the main reasons for QUIC is to allow it to evolve quickly as TCP is almost impossible to change as it’s so baked in. Future versions of QUIC will likely include Forward Error Correction (to automatically recreate dropped packets), connection migration (to allow you to seamlessly switch from WiFi to Mobile) and also be available for more than HTTP but they are out of scope for the initial version as defined by the QUIC working group charter because even without those QUIC is complicated. Additionally TCP is highly optimised into Operating Systems and network stacks so QUIC will likely be more CPU expensive and slow, especially initially and there may be other issues to solve as well.
So all in all, if you want QUIC now then look at one of the webservers or CDN or Google Cloud Platform and put this in front of your application server. Like HTTP/2 this usually gives the main benefits, and means you don’t need to worry about all of the above complications. But to me, QUIC is one to watch for the future, and not something id want to turn in for now.
If interested in learning more about HTTP/2, HTTP/3 and QUIC, and some of the complexities, then you can check out my book on the subject: https://www.manning.com/books/http2-in-action