n8n-nodes-reranker-voyageai

Developed and maintained by Jay Nguyen (Nguyễn Thiệu Toàn).

🛡️ Verified n8n Creator | 💼 CEO/Founder of GenStaff

Connect with me:

Voyage AI Reranker node for n8n. This node allows you to use Voyage AI's powerful reranking models (e.g. rerank-2.5) to reorder retrieved documents from a vector store, improving search relevance in RAG architectures.

Features

• Advanced Reranking: Uses Voyage AI's Cross-Encoder architecture to precisely evaluate document relevance.
• Top K Selection: Filter down to only the most relevant documents to save LLM token costs.
• Instruction Following: Compatible with rerank-2.5 which can follow natural language instructions in the query.
• Zero-dependency: Built natively for n8n without heavy external SDK dependencies.

Installation

Go to Settings > Community Nodes in your n8n instance and install:

n8n-nodes-reranker-voyageai

Credentials Configuration

1. Get your API Key from the Voyage AI Dashboard.
2. In n8n, set up a new Voyage API credential:
   • API Key: Enter your Voyage API key.

Usage

This node is meant to be used as a Sub-node (AI Reranker) connected to a Document Retriever (such as the Vector Store Retriever in n8n's Advanced AI tools).

1. Add a Vector Store node and set its operation to Retrieve.
2. Connect the Voyage AI Reranker node to the Document Compressor / Reranker input of the Vector Store node.
3. Configure the Model (e.g. rerank-2.5-lite) and Top K documents to return.

Workflow Example

(Provide a JSON workflow snippet here so users can copy-paste it directly into their n8n canvas)

{
  "nodes": [
    {
      "parameters": {
        "model": "rerank-2.5-lite",
        "topK": 3
      },
      "id": "example-voyage-reranker",
      "name": "Voyage AI Reranker",
      "type": "n8n-nodes-reranker-voyageai.rerankerVoyageAi",
      "typeVersion": 1,
      "position": [400, 200]
    }
  ],
  "connections": {}
}

Nodes

License

MIT