GPT-6 Sol feels weaker than 5.6 Sol on real coding work, critic says
A post criticizing OpenAI says GPT-6 Sol feels weaker than 5.6 Sol on many real coding tasks and Luna is noticeably dumb, while also complaining that quotas burn faster and resets are scarcer.
it's looking awkward for OpenAi > they're behind in intelligence compared to opus 5.5 now > compute looking rekt, quotas burning way faster than before > they got everyone addicted to resets like heroin and now they're not giving many > astra hype didn't last 2 weeks once the real cost showed up > gpt-6 sol feels weaker than 5.6 sol on a lot of real coding work > luna is awkwardly dumb even on the grunt work it's made for i've been hating on anthropic for like 3 months but ngl they deserve the credit on this one… let's just hope they don't nerf it anytime soon and openai better come up with something interesting, otherwise i don't see anyone staying on codex next month imo