Jina Deepsearch V1

jina-reranker-v3.5 is a 0.6B-parameter multilingual listwise document reranker and a drop-in upgrade to jina-reranker-v3. It keeps the same last-but-not-late (LBNL) listwise interface — ranking a query against many candidates in one forward pass — while improving domain robustness, structured-data ranking, multilingual retrieval, and inference efficiency.

A 3.8-billion-parameter general vector model (embedding model) for state-of-the-art multilingual embeddings for edge deployment.

A 3.8-billion-parameter general vector (embedding) model providing state-of-the-art multilingual embeddings with task-specific adapters.

A general-purpose vector model with 3.8 billion parameters, used for multimodal and multilingual retrieval, supporting both unidirectional and multi-vector embedding outputs.

Multimodal multilingual document reranker, 131K context, 0.6B parameters, for visual document sorting.

Multi-modal Embeddings Model, multilingual, 1024-dimensional, 865M parameters.