Chonkie is an artificial intelligence company that develops open-source context-engineering libraries used to prepare data for retrieval-augmented generation and large language model applications. Its flagship library splits large documents into semantically meaningful chunks, alongside companion tools for high-throughput chunking, embeddings clients, and retrieval-quality benchmarking.
The company also offers a model-specialization service in which its team trains custom models on a customer's production data so that software workflows run faster, more cheaply, and in closer alignment with business rules. Its open-source libraries are distributed under a permissive license, complemented by a sales-led engagement for teams that want specialized models.
Open-source Python libraries for cleaning, chunking, embedding, and storing data for retrieval-augmented generation and large language model pipelines.
