New: Qwen4-Max eval report is here.Read the report ↗

Best models for real-world agent tasks

AgentArena

Real-world agent tasks. Published Qwen4-Max launch results.

Updated: Data as of:
Limited reportingSaturation: not applicable
Official benchmark
AgentArena results. Each row identifies its benchmark version and evaluation track. Scores from different versions are not directly comparable. Best effort from the latest stored report date per model and version.
Model / creatorVersionReported score
Qwen4-MaxQwen · ProprietaryQwen4 launch snapshot7.5 points
View all 1 published reports