Mxbai Reranker Large V1

mixedbread-ai/mxbai-rerank-large-v1 is a reranking model that reorders candidate documents based on their relevance to a given query, improving the precision of retrieval systems.

This large-scale transformer model specializes in the reranking stage of retrieval pipelines. It accepts a list of query-document pairs and outputs a relevance score for each, enabling more accurate ranking than first-stage retrieval alone. The model is designed to capture fine-grained semantic relationships between queries and documents, making it suitable for applications such as search, question answering, and recommendation systems that require high relevance discrimination.

Gigarouter hosts mixedbread-ai/mxbai-rerank-large-v1 as a managed API with OpenAI-compatible endpoints, allowing developers to integrate reranking capabilities without managing infrastructure.