DeepSeek-V4.1-Flash Ranks Third Among Open Models on Agent Arena

Agent Arena launched a dynamic leaderboard ranking models by task completion and cost per task. DeepSeek-V4.1-Flash (Max) ranks third among open models (12th overall) with a +4.87% net improvement at roughly $0.06-0.07 per task, placing it on the Pareto frontier.

2026-09-15 ~ 2026-09-15 · 3 related posts