https://store-images.s-microsoft.com/image/apps.57941.29884b11-fcd2-4976-8552-c22b4dd6d1e8.bf88dd75-e00c-4b48-9fc1-0a188168848a.987bfd2c-b75a-4f53-a53c-19de170fe150

voyage-law-2 Embedding Model

Voyage AI Innovations Inc

voyage-law-2 Embedding Model

Voyage AI Innovations Inc

Text embedding model optimized for legal retrieval and AI applications. 16K context length.

Text embedding models are neural networks that transform texts into numerical vectors. They are a crucial building block for semantic search/retrieval systems and retrieval-augmented generation (RAG) and are responsible for the retrieval quality. voyage-law-2 is a cutting-edge embedding model that is optimized for semantic retrieval of legal texts. The model excels in law-related AI applications, including semantic case retrieval, legal question answering, and various functions of general legal AI assistants. On 8 legal retrieval tasks, voyage-law-2 has a significant 5.62% improvement over any alternatives, including OpenAI v3 large, Cohere English v3 and E5 Mistral. voyage-law-2 also has consistent enhancements across general-purpose corpora and long-context retrieval tasks, exceeding OpenAI v3 large on average by over 15%. Learn more about voyage-law-2 here.