Show HN: I reduced LLM inference GPU calls by 94% using semantic routing

(icomnewtechnologies.com) by kanacki | Jun 1, 2026 | 0 comments on HN