The ubiquitous occurrence of cluster patterns in nature still lacks a comprehensive understanding. It is known that the dynamics of many such natural systems is captured by ensembles of Stuart-Landau oscillators. Here, we investigate clustering dynamics in a mean-coupled ensemble of such limit-cycle oscillators. In particular we show how clustering occurs in minimal networks, and elaborate how the observed 2-cluster states crowd when increasing the number of oscillators. Using persistence, we discuss how this crowding leads to a continuous transition from balanced cluster states to synchronized solutions via the intermediate unbalanced 2-cluster states. These cascade-like transitions emerge from what we call a cluster singularity. At this codimension-2 point, the bifurcations of all 2-cluster states collapse and the stable balanced cluster state bifurcates into the synchronized solution supercritically. We confirm our results using numerical simulations, and discuss how our conclusions apply to spatially extended systems.Certain swarms of fireflies are known to flash in unison. They also sometimes divide into two or more distinct yet internally synchronized groups, flashing with a certain phase lag between the groups. This is just one example of clustering dynamics in an ensemble of coupled oscillators, as it occurs naturally in many physical systems. A key problem in the understanding of clustering dynamics is the connection between its occurrence in small and large ensembles. In other words, is there a universal law governing the arrangement of cluster states, independent of the system size? This paper partially answers this question and links the phenomenon of clustering in minimal networks of globally coupled limit-cycle oscillators to clustering in ensembles of infinitely many oscillators. We demonstrate that a natural arrangement of such 2-cluster states exists: When tuning a parameter, a balanced cluster state transitions to synchronized motion via a sequence of intermediate unbalanced cluster states. Tuning an additional parameter, this sequence converges to a single point in parameter space where all cluster states are born directly at the synchronized solution. We call such a codimension-2 point a cluster singularity. Singularities of this kind may appear in any symmetrically coupled ensemble of oscillators, and thus play a crucial role for the understanding of collective behavior in oscillatory systems.Clustering, as discussed above, typically occurs in systems with long-range interactions 1 .Oscillatory systems with global interactions play a crucial role in the understanding of phenomena observed in nature and technology, such as the visual perception in the mammalian brain, the circadian rhythm in the heart or the behavior of coupled Josephson junctions and electrochemical oscillators. Even phenomena such as synchronous chirping of crickets, the flashing of fireflies in unison and the synchronous clapping of an audience can be traced back to the action of a long-range coupling between oscillating un...