Skip to content

Comment on GPT-5.2 and GPT-5.2-Codex are now 40% fasterparent

Comments

This makes no sense. Why lower extended thinking time? Those who want faster answers can just use standard. The only purpose this serves is to "trick" the user into thinking he's still receiving "extended thinking" level answers at faster speed.

AboutSource Built by g1lg1l

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