Skip to content

Comment on Ask HN: Does anyone else find GPT-5.6 Sol in Codex slow?

Comments

It feels like all models are slow(er). I've switched back to 5.5 Medium and it also took 12 minutes to do a small-ish refactor across 8 files. I guess that happens when you onboard nearly a million users per day while also removing the 5h usage limit.

AboutSource Built by g1lg1l

Hackerly is an independent reader for Hacker News, built on the public HN API. Not affiliated with Y Combinator.