Comment

Comment by ashvardanian

Which models would be that expensive to run? Or is there like a 1000x amplification for each prompt?

Discussion

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton