Snowflake Arctic Embed is a family of open-source text embedding models, licensed under Apache 2.0, specifically designed for retrieval use cases. These models offer leading retrieval performance on the Massive Text Embedding Benchmark (MTEB) and provide organizations with new advantages when combining proprietary datasets with Large Language Models (LLMs) for Retrieval Augmented Generation (RAG) or semantic search services. They come in various sizes, from extra-small (xs) to large (l), with different context windows and parameter counts to meet the diverse latency, cost, and retrieval performance needs of businesses.