Jina Reranker v2 Base Multilingual
Jina Reranker v2 scores a query together with a document and orders candidate results by relevance. Its training also covers code retrieval, function-related queries and text-to-SQL retrieval.
How the model works
The model card documents a 1,024-token scoring window and a sliding-window method for longer input. Its rerank function combines scores across chunks. This distinction matters because other provider material lists longer document support: accepting a long document through chunking is not the same as scoring every token in one native window.Jina Reranker v2 Base Multilingual model details
Use and access
The release provides compute_score and rerank interfaces, plus Transformers and JavaScript examples. Full weights are licensed under CC BY-NC 4.0, with separate managed commercial access through Jina and cloud marketplaces. It is a text ranker; m0 adds document-image understanding and v3 introduces joint listwise processing.Jina Reranker v2 Base Multilingual model detailsJina Reranker v2 Base Multilingual technical overview
Specifications & access
| Detail | Value and conditions |
|---|---|
| Launch date | |
| Access | API, Downloadable |
| Model identifiers | jina-reranker-v2-base-multilingualjinaai/jina-reranker-v2-base-multilingual |
| Context window | 1,024 tokens Native scoring window stated in this model card; longer documents are processed in overlapping chunks. Jina Reranker v2 Base Multilingual model details |
| Parameters | Approximately 278 millionJina Reranker v2 Base Multilingual model details |
| Weight license | CC BY-NC 4.0Jina Reranker v2 Base Multilingual model details |
Related models
- Jina Reranker v3
Jina AI · Jina Reranker
A compact ranker that compares a query with several documents in one pass.
Launch date API · Downloadable - Jina Reranker m0
Jina AI · Jina Reranker
A relevance scorer for text and document images in multiple languages.
Launch date API · Downloadable