Announcement
Claude Opus 5.5 (Max) tops Agent Arena’s updated leaderboard
Agent Arena says its Bash Reliability signal evolved from Bash Recovery, while proprietary models occupy the top 14 spots.
TLDR
Agent Arena says it evolved its Bash Recovery signal into Bash Reliability and updated its leaderboard. Claude Opus 5.5 (Max) ranks first, followed by GPT-6.1 Sol (Max), Claude Fable 5.1 (Max), GPT-6 Astra (Max) and Gemini 4 Argon (High). It says proprietary models occupy the first 14 places; GLM 5.2 at No. 15 and MiMo V2.6 Pro at No. 18 are the highest-ranked open-weights models.
Combined views
—
3 Sources, first seen ago
— likes— comments— saves— reposts