SPARQL query services that balance processing between clients andservers become more and more essential to endure the increasing load on published open and decentralized knowledge graphs overthe Web. To this end, Linked Data Fragments (LDF) have introduceda foundational framework that has sparked research exploring aspectrum of potential Web querying interfaces in between SPARQL endpoint servers on the one end, and client-side processing of datadumps on the other. Current proposals in between usually suffer from imbalanced load on either the client (TPF, smart-KG) or theserver (SaGe, SPF) side. The present paper, to the best of our knowledge, is the first work to combine both client-side and server-side query processing optimizations in a truly dynamic fashion: we introduce WiseKG, which employs a cost model that dynamically delegates the load between servers and clients by combining client-side processing of shipped partitions (á la smart-KG) with efficientserver-side processing of star-shaped sub-queries (á la SPF), based on current server workload and client capabilities. Our experiments show that WiseKG significantly outperforms state-of-the-art so- lutions in terms of average total query execution time per client, while at the same time decreasing network traffic, and increasing server-side availability.