OpenAI-3072

OpenAI-3072 refers to the high-dimension output of the text-embedding-3-large model designed for complex semantic search and RAG applications.

This configuration leverages the text-embedding-3-large model to generate 3,024-dimensional vectors (often shortened to 3072 in documentation) for superior information retrieval. By mapping text to a massive coordinate space, it captures nuanced linguistic relationships that smaller models miss. It currently costs $0.00013 per 1k tokens, making it a cost-effective choice for developers building enterprise-grade search engines, recommendation systems, or long-context memory for LLMs.

https://platform.openai.com/docs/guides/embeddings