There is a need to break long data into segments known as chunks for LLM vector databases due to model data processing limitations. Chunking allows to segment data to fit the limitations of large language models, enabling effective context preservation, system efficiency and cost control. Various chunking strategies can be used depending on the nature of the content embedded, the model for embedding, the expectations for user queries, context window limitations and other parameters of LLM applications. Small chunks focus on individual sentence's meaning while larger chunks capture broader meaning and context. Fixed-size vs. context-aware chunking strategies exist, with fixed-size using a predetermined number of characters or words to segment text, and context-aware splitting text into segments using context separators such as periods. Chunk overlap preserves context between segments when using fixed-size chunking, mitigating context loss and maintaining LLM understanding of the whole text.