Jina-reranker-v3 is a multilingual document reranking model with 0.6B parameters that introduces a novel last-but-not-late interaction. Unlike late interaction models like ColBERT, it performs causal self-attention between the query and documents within the same context window, enabling rich cross-document interactions before extracting contextual embeddings from the last token of each document. This compact architecture achieves state-of-the-art BEIR performance with 61.94 nDCG@10, while being significantly smaller than generative list-wise reranking models.
