Command Code AI Hailed as "GOAT" After Topping Composio Benchmark

Key Info

Command Code AI has been hailed as the "GOAT" after topping Composio's benchmark for knowledge work, outperforming every other harness on open models like DeepSeek with 18/30 tasks solved and the fastest average time of 375 seconds.

Highlights

• Solved 18/30 tasks, the most accurate among tested harnesses on open models.
• Fastest average task completion time at 375 seconds, beating even DeepSeek's own harness.
• Also recognized as one of the most token-efficient top-tier harnesses, leading in accuracy, cost, and speed across both open and closed frontier models.