This is similar to a runtime swap: same UI, different inference backend

→ View original post on X — @nicos_ai