It's wild how underrated the new OpenAI codex models are!

And can I say - 10x rate limits vs opus 4.6

@VaibhavSisinty It ships production-grade diffs while others polish prompts. Underrated for builders who ship.

@VaibhavSisinty What feels most underrated to you right now — reliability, speed, or just how usable they are in actual flows?

@VaibhavSisinty Codex models are quietly becoming very capable, especially when used inside structured workflows instead of one-off prompts. The gap now is not capability, it is how well people design systems around them.

@VaibhavSisinty Me switching between Code tools depending on X vibes.

@VaibhavSisinty People are just stuck to Claude Code atm

@VaibhavSisinty Are they good enough? I m busy with claude anyways

@VaibhavSisinty timeline differences. my tl is filled with people preaching gpt 5.4 xhigh. these models have made me very bullish on oai

@VaibhavSisinty Yea!! Everyone keeps talking about Claude code, which is great, but codex IMO is better!

@VaibhavSisinty There’s no going back to open AI. Hooked to Claude

@VaibhavSisinty I have switched to using Codex for almost everything. It does even non coding tasks better than anything available right now.

@VaibhavSisinty most people no need superintelligence for his

@VaibhavSisinty OpenAi model is upgrading really well and fast