Skip to main content

Vector

The vector field type is used to store high-dimensional embeddings, typically generated from text, files, or chat conversations. These embeddings are numerical representations of content that enable powerful applications like semantic search, similarity matching, clustering, and classification.

Capabilities Overview​

The vector field is a cornerstone of AI-driven workflows. You can generate embeddings from various sources including raw text, chat fields, and files using models provided by OpenAI and Unstructured. These embeddings can then be stored in the vector field for downstream operations such as similarity comparisons or searching within a vector database.

Embedding generation is supported through multiple operations:

  • Text and chat embeddings using OpenAI’s models (like text-embedding-3-small or text-embedding-ada-002)
  • File-based embeddings via chunking strategies and OCR processing
  • Multi-paragraph (chunked) embeddings to retain semantic structure

This field type is essential for building features like knowledge retrieval, AI search interfaces, document understanding, and content recommendation systems.