jina-reranker-v3.5-mlx

MLX port of jina-reranker-v3.5, a domain-ready multilingual listwise reranker optimized for Apple Silicon. No transformers library required.

Installation

pip install -r requirements.txt

Usage

from rerank import MLXReranker

reranker = MLXReranker()

query = "What are the health benefits of green tea?"
documents = [
    "Green tea contains antioxidants called catechins that may help reduce inflammation.",
    "Studies show that drinking green tea regularly can improve brain function.",
    "Basketball is one of the most popular sports in the United States.",
]

results = reranker.rerank(query, documents)
for result in results:
    print(f"Score: {result['relevance_score']:.4f}  {result['document'][:80]}...")

Returns: List of dicts sorted by relevance descending — document, relevance_score, index, and optionally embedding (shape (512,)).

reranker.rerank(
    query: str,
    documents: List[str],
    top_n: Optional[int] = None,
    return_embeddings: bool = False,
)

Custom Model Path

reranker = MLXReranker(
    model_path="/path/to/jina-reranker-v3.5-mlx",
    projector_path="/path/to/jina-reranker-v3.5-mlx/projector.safetensors",
)

Citation

If you find jina-reranker-v3.5 useful in your research, please cite the original paper:

@misc{nasika2026jinarerankerv35efficientlistwisereranker,
      title={jina-reranker-v3.5: An Efficient Listwise Reranker with Hybrid Attention and Self-Distillation}, 
      author={Christina Nasika and Feng Wang and Antonis Krasakis and Han Xiao},
      year={2026},
      eprint={2607.18152},
      archivePrefix={arXiv},
      primaryClass={cs.IR},
      url={https://arxiv.org/abs/2607.18152}, 
}