Voyage Rerank 2.5

Voyage Rerank 2.5 is a reranking model developed by Voyage AI, designed to reorder candidate documents retrieved by a first-stage retrieval system so that the most relevant results appear at the top. It accepts a query and a list of documents, then assigns relevance scores that can be used to re-sort results before passing them to a downstream application. With a 32,000-token context window, it can process longer documents without truncation, which is useful in retrieval-augmented generation (RAG) pipelines where document chunks may be substantial.

Reranking models like Voyage Rerank 2.5 are typically used as a second-stage component after an initial embedding-based or keyword search retrieval step. By applying a more precise relevance judgment at this stage, they help reduce noise in the retrieved set before it reaches a language model or end user. This makes the model well-suited for search applications, question-answering systems, and any workflow where retrieval precision directly affects output quality.