The task of Entity Summarization (ES) focuses on capturing the most relevant and representative information about an entity, facilitating quick exploration and understanding. In this demonstration, we present EntitySum, an approach that combines centrality measures for object weighting, and frequency-based methods for property weighting in order to select the most appropriate triples to be included in the Entity Summary. The process filters out irrelevant data, reduces redundancy, and generates concise, informative entity summaries. As a next step, large language models are utilized to transform the Entity Summaries into human-readable text. During the demonstration, we will enable conference participants to actively use EntitySum on DBpedia and assess the quality of the resulting Entity Summaries. Further, we will allow users to contrast the textual representation of various large language models (LLM) and identify the impact of the various components.