Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

this comparison for the quantization effect is very nice https://flux-quality-comparison.vercel.app/

however i do have to ask.. ~2x faster for fp16->fp8 is expected right? its still not as good as the "realtime" or "lightning" options that basically have to be 5-10x faster. whats the ideal product usecase for just ~2x faster?



Funny, sometime I like the fast one better.




Consider applying for YC's Winter 2026 batch! Applications are open till Nov 10

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: