OpenAI embedding model

OpenAI embedding models convert text into numerical vectors (embeddings), quantifying semantic relatedness for tasks like search and clustering.

These are high-performance models, including the latest `text-embedding-3-large` and the widely adopted `text-embedding-ada-002`, that transform text into dense vector representations. The vectors capture semantic meaning: closer vectors in the high-dimensional space mean greater similarity between text strings. This capability is foundational for applications like semantic search, content clustering, and recommendation systems. For example, the `text-embedding-3-small` model offers a 1536-dimensional vector output, while the `large` model provides a 3072-dimension vector, offering superior performance on benchmarks like MTEB and MIRACL.

https://platform.openai.com/docs/guides/embeddings/what-are-embeddings