1: \begin{abstract}

2: We study the complexity of producing $(\delta,\epsilon)$-stationary points of Lipschitz objectives which are possibly neither smooth nor convex, using only noisy function evaluations.

3: Recent works proposed several stochastic zero-order algorithms that solve this task, all of which suffer from a dimension-dependence of $\Omega(d^{3/2})$ where $d$ is the dimension of the problem, which was conjectured to be optimal. We refute this conjecture by providing a faster algorithm that has complexity $O(d\delta^{-1}\epsilon^{-3})$, which is optimal (up to numerical constants) with respect to $d$ and also optimal with respect to the accuracy parameters $\delta,\epsilon$, thus solving an open question due to \citet{lin2022gradient}. Moreover, the convergence rate achieved by our algorithm

4: is also optimal for smooth objectives,

5: proving that in the nonconvex stochastic zero-order setting, \emph{nonsmooth optimization is as easy as smooth optimization}. 

6: We provide algorithms that achieve the aforementioned convergence rate in expectation as well as with high probability.

7: Our analysis is based on a simple yet powerful lemma regarding the Goldstein-subdifferential set,

8: which allows utilizing recent advancements in first-order nonsmooth nonconvex optimization.

9: \end{abstract}

10: