Qdrant Vector Store

​

Node Details

• Name: Qdrant_VectorStores

• Type: Qdrant

• Version: 5.0

• Category: Vector Stores

​

Base Classes

• Qdrant

• VectorStoreRetriever

• BaseRetriever

​

Input Parameters

​

Main Parameters

1. Document (optional, list)

   • Type: Document

   • Description: List of documents to be processed

2. Embeddings

   • Type: Embeddings

   • Description: Embedding model to use

3. Record Manager (optional)

   • Type: RecordManager

   • Description: Keeps track of records to prevent duplication

4. Qdrant Server URL

   • Type: string

   • Placeholder: http://localhost:6333

5. Qdrant Collection Name

   • Type: string

6. File Upload (optional)

   • Type: boolean

   • Description: Allows file upload on the chat

​

Additional Parameters

1. Vector Dimension

   • Type: number

   • Default: 1536

2. Content Key (optional)

   • Type: string

   • Default: ‘content’

   • Description: Key for storing text

3. Metadata Key (optional)

   • Type: string

   • Default: ‘metadata’

   • Description: Key for storing metadata

4. Upsert Batch Size (optional)

   • Type: number

   • Description: Upsert in batches of size N

5. Similarity

   • Type: options

   • Options: Cosine, Euclid, Dot

   • Default: Cosine

6. Additional Collection Configuration (optional)

   • Type: json

7. Top K (optional)

   • Type: number

   • Description: Number of top results to fetch (default: 4)

8. Qdrant Search Filter (optional)

   • Type: json

   • Description: Conditions for returning points

​

Outputs

1. Qdrant Retriever

   • Type: retriever

   • Base Classes: [Qdrant, VectorStoreRetriever, BaseRetriever]

2. Qdrant Vector Store

   • Type: vectorStore

   • Base Classes: [Qdrant, QdrantVectorStore base classes]

​

Credentials

• Credential Name: qdrantApi

• Parameters: qdrantApiKey

​

Functionality

• Supports upsert operations for adding or updating documents in the Qdrant collection

• Enables similarity search and retrieval of documents

• Allows for batch processing of documents

• Supports custom configuration for Qdrant collections

• Integrates with a record manager to prevent duplication

• Provides options for different similarity measures (Cosine, Euclid, Dot)

• Supports file upload functionality in chat interfaces

​

Use Cases

• Semantic search applications

• Document retrieval systems

• Recommendation engines

• Any application requiring efficient storage and querying of vector embeddings

​

Note

This node is designed to work with the Qdrant vector database and requires proper setup of the Qdrant server. It’s particularly useful for applications dealing with large-scale vector data and requiring fast similarity searches.