Oracle AI Database Embeddings
Use the n8n Oracle AI Database Embeddings sub-node to generate embeddings with an ONNX embedding model loaded in Oracle AI Database.
Contents
Requirements
Load an ONNX embedding model into Oracle AI Database before you use this node. The Oracle user used by n8n must be able to invoke the model and perform the required vector operations.
The embedding model’s output dimension must match the vector column dimension used by the Oracle AI Database Vector Store node.
Configure the Embeddings Node
Add the Embeddings Oracle AI Database node as an AI cluster sub-node. Select or create an Oracle AI Database credential. The node uses the database-resident ONNX model to create embeddings; it does not require a separate external embedding service.
The node is intended for use with the Oracle AI Database Vector Store node. Select the model configured in Oracle AI Database when the node exposes the model field.
Connect to a Vector Store
Connect the Embeddings Oracle AI Database node to the embeddings connector of the Oracle AI Database Vector Store node. Use the vector store to insert documents or retrieve documents for an AI chain, retriever, question-answer tool, or AI agent.
Keep the same embedding model and dimension for document insertion and query retrieval. Changing either value can make stored vectors incompatible with new query vectors.