Pigeonhole and Extremal Arguments · Challenge problem

Claim: every positive integer $n$ with $n \equiv 1 \pmod 4$ is a sum of two squares ($0$ allowed). What is the smallest counterexample?