• Jina Reranker m0 is a neural multimodal reranking model, designed to enhance the relevance of search results.
• It complements text, image or multimodal embedding models and refines search results by prioritizing documents relevant to a query.
• This state-of-the-art reranker model enables a variety of applications that rely on precise search results across different languages, improved information retrieval, and high document throughput. Use-cases: Deep search, AI agents orchestration, vector search, retrieval augmented generation (RAG).

Highlights of Jina Reranker m0

• High multilingual performance across the board: This reranker model ranks at the top compared to its competitors, in terms of NDCG and recall, according to MBEIR, MKQA, ViDoRe and others.
• Extended context length: This reranker model is capable of handling queries up to 10,240 tokens.
• Novel support for image documents: One of the first reranker models to support text, images and complex documents with negligible modality gap.