Report
Claude Sonnet 5.5 (Max) debuts at No. 3 in Agent Arena
Agent Arena says the model leads its Chat category with +15.6% net improvement, but costs a median $2.74 per task.
TLDR
Agent Arena says Anthropic’s Claude Sonnet 5.5 (Max) debuted at No. 3 with a +12.5% net improvement score, up 8.1 percentage points from Sonnet 5 (High). It ranked No. 1 in Chat, but its median cost of $2.74 per task was about 73% higher than No. 2 Claude Opus 5.5 (High), which Agent Arena says scored higher.
Combined views
49.1K
6 Sources, first seen ago
