JinaRanker

Re-rank documents based on their relevance to a query using Jina AI reranker models.

Key Features

• Re-ranks documents by relevance to a query using Jina AI cross-encoder models.
• Improves retrieval quality by directly comparing query-document pairs.
• Configurable top_k to limit the number of returned documents.
• Configurable score_threshold to filter out low-relevance documents.

Configuration

1. Drag the JinaRanker component onto the canvas from the Component Library.
2. Click on the component to open the configuration panel.
3. On the General tab:
   1. Create a secret with your Jina API key. Use JINA_API_KEY as the secret key. For instructions, see Create Secrets. Get your API key from Jina AI.
   2. Select the reranker model. Check available models on the Jina documentation.
   3. Set top_k to control the number of documents returned.
4. Go to the Advanced tab to configure score_threshold.

Connections

JinaRanker accepts a query string through its query input, a list of documents through its documents input, and optional top_k and score_threshold overrides at runtime. It outputs ranked documents through its documents output, sorted from most to least relevant.

Connect a Retriever's documents output to JinaRanker's documents input. Then connect JinaRanker's documents output to PromptBuilder or another component that uses ranked documents.

Source Code

To check this component's source code, open ranker.py in the Haystack Core Integrations repository.

Usage Examples

Basic Configuration

  JinaRanker:
    type: haystack_integrations.components.rankers.jina.ranker.JinaRanker
    init_parameters:
      api_key:
        type: env_var
        env_vars:
        - JINA_API_KEY
        strict: false
      model: jina-reranker-v1-base-en
      top_k: 3

This example shows a RAG pipeline that retrieves documents, re-ranks them using Jina, and generates an answer.

components:
  InMemoryBM25Retriever:
    type: haystack.components.retrievers.in_memory.bm25_retriever.InMemoryBM25Retriever
    init_parameters:
      document_store:
        type: haystack.document_stores.in_memory.document_store.InMemoryDocumentStore
        init_parameters:
          bm25_tokenization_regex: (?u)\b\w\w+\b
          bm25_algorithm: BM25L
          bm25_parameters:
          embedding_similarity_function: dot_product
          index: 'default'
          async_executor:
      top_k: 10
  JinaRanker:
    type: haystack_integrations.components.rankers.jina.ranker.JinaRanker
    init_parameters:
      api_key:
        type: env_var
        env_vars:
        - JINA_API_KEY
        strict: false
      model: jina-reranker-v1-base-en
      top_k: 3
  PromptBuilder:
    type: haystack.components.builders.prompt_builder.PromptBuilder
    init_parameters:
      template: "Given the following documents, answer the question.\n\nDocuments:\n{% for doc in documents %}{{ doc.content }}\n{% endfor %}\n\nQuestion: {{ query }}"
  OpenAIGenerator:
    type: haystack.components.generators.openai.OpenAIGenerator
    init_parameters:
      api_key:
        type: env_var
        env_vars:
        - OPENAI_API_KEY
        strict: false
      model: gpt-4o-mini
  AnswerBuilder:
    type: haystack.components.builders.answer_builder.AnswerBuilder
    init_parameters:
      pattern:
      reference_pattern:
      last_message_only: false
      return_only_referenced_documents: true

connections:
- sender: InMemoryBM25Retriever.documents
  receiver: JinaRanker.documents
- sender: JinaRanker.documents
  receiver: PromptBuilder.documents
- sender: PromptBuilder.prompt
  receiver: OpenAIGenerator.prompt
- sender: OpenAIGenerator.replies
  receiver: AnswerBuilder.replies
- sender: JinaRanker.documents
  receiver: AnswerBuilder.documents

max_runs_per_component: 100

metadata: {}

inputs:
  query:
  - InMemoryBM25Retriever.query
  - JinaRanker.query
  - PromptBuilder.query
  - AnswerBuilder.query

outputs:
  answers: AnswerBuilder.answers

Parameters

Inputs

Parameter	Type	Description
query	str	Query string.
documents	List[Document]	List of Documents.
top_k	Optional[int]	The maximum number of Documents you want the Ranker to return.
score_threshold	Optional[float]	If provided, only returns documents with a score above this threshold.

Outputs

Parameter	Type	Description
documents	List[Document]	List of Documents most similar to the given query in descending order of similarity.

Init Parameters

These are the parameters you can configure in Pipeline Builder:

Parameter	Type	Default	Description
api_key	Secret	Secret.from_env_var('JINA_API_KEY')	The Jina API key. It can be explicitly provided or automatically read from the environment variable JINA_API_KEY (recommended).
model	str	jina-reranker-v1-base-en	The name of the Jina model to use. Check the list of available models on https://jina.ai/reranker/
top_k	Optional[int]	None	The maximum number of Documents to return per query. If None, all documents are returned
score_threshold	Optional[float]	None	If provided only returns documents with a score above this threshold.

Run Method Parameters

These are the parameters you can configure for the component's run() method. This means you can pass these parameters at query time through the API, in Playground, or when running a job. For details, see Modify Pipeline Parameters at Query Time.

Parameter	Type	Default	Description
query	str		Query string.
documents	List[Document]		List of Documents.
top_k	Optional[int]	None	The maximum number of Documents you want the Ranker to return.
score_threshold	Optional[float]	None	If provided only returns documents with a score above this threshold.