Embedditor is an open-source tool that helps users improve the performance of their vector search capabilities by enhancing embedding metadata and tokens. It offers a user-friendly interface to apply advanced NLP cleansing techniques, split or merge content intelligently, and add void or hidden tokens to make chunks more semantically coherent. Embedditor can be deployed locally on a PC or in a dedicated cloud or on-premises environment, providing full control over data. By filtering out irrelevant tokens, users can save up to 40% on the cost of embedding and vector storage while getting better search results.