Open Agent Leaderboard Launched for General-Purpose AI Agents

A new initiative called the Open Agent Leaderboard has been launched to evaluate and compare full AI agent systems. This leaderboard, paired with the Exgentic framework, assesses agents across six realistic task benchmarks including coding, customer service, and research, measuring both quality and cost. Remarkably, the initial findings show that general-purpose agents can compete with specialized ones, sometimes matching systems built directly for specific tasks.