Is Opus 5.5 really dumber? Complaints flood in over claim that "something changed on Nov. 3"

Claims that Anthropic's top model, Claude Opus 5.5, has gotten worse since launch have spread on X for days. As of Monday morning, Anthropic had issued no public statement confirming or rejecting them.

Shane Melaugh posted on Nov. 7 that he had Gemini 4 Argon analyze 1,400 Claude Code sessions and "on November 3, something definitely happened. The model name didn't change, but the way it does reasoning took a sharp turn." The post drew about 2,400 likes and 380,000 views. In the same thread he said a bug that "high" effort could not fix in three hours was fixed by xhigh in minutes, and that the "use high for everything" rule "no longer applies." Developer Theo replied that he had told a bot to check the claims against his own logs, and separately posted that he was "going back to Opus 4.5 just to feel something" (about 2,300 likes).

Complaints also turned to pricing. AIMind posted two videos from the same prompt on Nov. 8, saying the $20 plan was enough in September but he now pays $100 (about 2,800 likes). A reply countered that running one prompt once each is not a controlled test.

Outside trackers have not caught a drop. The open-source LiveNerf, which runs a frozen 78-question panel daily since launch, reported "no change" in all three 10-day windows as of Nov. 7, though it says the smallest shift it can detect per window is about 6.6 points. Anthropic's docs list Opus 5.5's default effort as medium, below Opus 5's high, so requests that don't set effort run a step lower than before. Details are in the feature below.