Voyage embeddings are text embedding models from Voyage AI, offered through an API, with general-purpose versions and special ones for code, finance and law.
In plain words
Voyage models turn text into vectors that capture meaning, mainly for search and RAG. The voyage-4 series has large, standard and lite models whose vectors are compatible with each other, so you can mix them. They read up to 32,000 tokens and let you choose vector sizes from 256 to 2,048 numbers. For search, you tell the API whether each text is a query or a document.
A quick example
A legal team building a contract search tool could embed every clause with voyage-law-2, then embed each lawyer's question as a query and return the clauses with the closest vectors.
Where you'll meet it
- Retrieval for RAG and semantic search.
- Domain search over code, financial or legal documents.
Related
Sources
- DocsText Embeddings, Voyage AI Docs · read 30 Sept 2026
- DocsRerankers, Voyage AI Docs · read 29 Sept 2026
- DocsThe Voyage 4 model family, Voyage AI · read 29 Sept 2026
Checked by WhatIsAI against the sources above · 29 Sept 2026. Spotted an error? Report it.