Beyond the Benchmark: The Binance Agentic AI Challenge

**Benchmarks are static. Reality isn't.** Build an LLM-powered Agentic AI system that gathers information, reasons, decides, manages its own risk, and operates reliably — then find out how it holds up over a 14-day run in a real-time environment. Fully audited. Fully traceable. The best system wins. # **Overview** **Applications**: Sep 9-30 — Submit your proposal through [this form](https://forms.gle/UmDXgdwPPwknYZur9) **Format**: Online **Duration**: 4 weeks of active work (2-week build + 2-week scored run); ~7 weeks end-to-end including onboarding and audit **Finalists**: