[How to: split text into semantic chunks](https://python.langchain.com/docs/how_to/semantic-chunker): LLM should read this page when building an application that needs to split long text into smaller chunks based on semantic meaning, when working with large documents that need to be broken down into semantically coherent sections, or when needing to control the granularity of text splitting. This page explains how to use the SemanticChunker from LangChain to split text into semantically coherent chunks by leveraging embedding models, with options to control the splitting behavior based on percentile, standard deviation, interquartile range, or gradient of embedding distance.

