Claude Opus 5.5 (High) ranks second in Arena.ai Agent Arena

Arena.ai ranks Claude Opus 5.5 (High) second in Agent Arena, with a net improvement score of +12.15%. Only Fable 5.1 (Max) ranks higher. At a median price of $1.31 per task, Arena.ai says Opus 5.5 (High) costs 64% less and moves the Pareto frontier.

The ranking also gives Opus 5.5 (High) a higher net improvement score than both earlier Opus 5 variants. Arena.ai says it costs 40% less than Opus 5 (High) and 56% less than Opus 5 (Max).

By signal, Opus 5.5 (High) ranks first in Steerability at +14.50%, second in Confirmed Success at +15.50%, third in Praise vs Complaint at +19.80%, and fourth in Bash Recovery at +10.64%.