← sources

even a cheaper model has to earn its place

source: reddit (reddit:SaaS)published 2026-10-11 14:40ingested 2026-10-12 02:37

retained text

How are people deciding whether a cheaper LLM is good enough for a feature? because saving on tokens makes total sense but less so if it takes three attempts to get a usable answer.
Are you testing it against your current model on real requests or just checking a few outputs before switching?

document fetched 2026-10-12 00:27 · canonical https://reddit.com/r/SaaS/comments/1x39wfb/even_a_cheaper_model_has_to_earn_its_place