OpenAI DevDay is Tuesday. How do you think Anthropic will crash the party?
retained text
OpenAI DevDay is on Tuesday. So how do you think Anthropic is going to crash the party? My guess is that we’ll see Sonnet 5.5 and Haiku 5.5 very soon. Anthropic has already confirmed that both are coming in the next few weeks, and the timing would be almost too perfect. And honestly, that could hurt. If they manage to bring the same kind of improvements we just saw with Opus 5.5 to Sonnet and Haiku, they could end up with models that are fast, cheap and actually really good. That’s exactly the combination OpenAI should probably be worried about right now. I don’t really see much urgency for a Fable 5.5 at the moment. Fable 5.1 is still extremely capable, and Opus 5.5 is already performing ridiculously well, often surprisingly close to it while being much cheaper. So now all we can do is hope OpenAI doesn’t spend DevDay showing us another collection of shiny toys that almost nobody asked for, and actually pulls a genuinely great model out of the hat. Because OpenAI increasingly feels like a runner who goes absolutely flat out for the first 300 meters, looks unstoppable, then runs out of oxygen in the final 100 while everyone else casually starts passing them. Sorry, Luna 6 and Sol 6 are pissing me off again. I still can’t believe what they did to those models. At this point, I’m far more interested in whether OpenAI can deliver a genuinely great model again than in another reset button, agent demo or flashy feature.
document fetched 2026-09-27 00:29 · canonical https://reddit.com/r/OpenAI/comments/1wqmp98/openai_devday_is_tuesday_how_do_you_think