Resources · n8n integration

n8n Embeddings Oracle Database nodeConfigure Embeddings Oracle Database in n8n.

The n8n Embeddings Oracle Database node turns text into vectors so another node can search on meaning. It is a sub-node, which means it has no main input and never runs alone. One parameter to fill, one Embeddings port to connect, and a root node that calls it.

Try the Embeddings Oracle Database node on n8nSee the 1 parameters

Verified Trustpilot reviews · AI, automation & growth agency

Why automate

What does the n8n Embeddings Oracle Database node actually do?

The node converts text into embeddings, that is lists of numbers that let a workflow compare meaning instead of matching exact words. Its n8n description says it in three words: use ONNX embeddings. It has no main input of its own. You drop it beside a root node, clip it onto the Embeddings port, and the root node calls it every time a piece of text has to become a vector.

First scenario, a searchable knowledge base. A vector store node sits in the workflow as the root node and needs something to embed the documents you push into it. This node fills that slot: pick the model in the Model field, connect the port, and every chunk that goes through the store gets its vector without any extra step in the canvas.

Second scenario, an assistant that answers from your own documents. An AI Agent queries a store to find the passages closest to a question. The same embedding node has to be attached on both sides, at indexing time and at query time, otherwise the numbers are not comparable and the search returns nothing useful.

Third scenario, sorting inbound text. Rows landing in Google Sheets or tickets arriving from a form can be embedded once, then compared to a handful of reference examples to spot near duplicates or route them to the right queue. The embedding sub-node does the numeric part, the root node does the logic.

When to prefer something else: this node runs ONNX embeddings. If the rest of your stack already relies on a hosted provider and its credential is set up, an embeddings sub-node from that provider, such as the one that pairs with OpenAI, keeps a single account to manage. The provider bills its own API calls in that case, and n8n adds nothing on top.

The limits are worth knowing before you build. The node exposes 1 parameter at version 1, so there is no batch size, no dimension override and no retry setting to tune here. It cannot be triggered, tested or run on its own: without a root node attached to its Embeddings port, it stays idle. And no public n8n template currently uses it, so the wiring is on you. The n8n training covers that sub-node model in practice.

Connect

What do you set up before the first run?

  1. 01

    Create the credential once

    Open the Credentials menu in n8n and create the credential this node asks for. A credential in n8n is a stored set of connection details, saved apart from the workflow so the values never sit in the canvas. It is created once and reused by every workflow on the instance, which matters as soon as the same embedding setup serves several automations.

  2. 02

    Select it in the node

    Open the Embeddings Oracle Database node and pick the credential from the dropdown at the top of the panel. n8n keeps the selection with the node, so duplicating the node into another workflow keeps the link. If the list is empty, the credential was saved under a different type and has to be created again from the node itself.

  3. 03

    Attach it to a root node

    A sub-node with a valid credential still does nothing on its own. Drag the connector from the Embeddings port to the matching input of the root node that needs vectors, then run the root node. The execution log shows the sub-node call nested under the root node, which is where you check that the credential actually worked.

Connections

What does this node connect to?

n8n splits its AI nodes in two: a root node receives the workflow items through its main input, and sub-nodes clip onto its ports, one type of sub-node per port. This node is a sub-node, and it offers a single port.

Output (what it plugs into)

  • Embeddingsai_embedding
01EmbeddingsThe Embeddings port is how this node hands its vectors to whatever asked for them. It is an output: the connection starts here and lands on the root node, not the other way round.

The Embeddings port is how this node hands its vectors to whatever asked for them. It is an output: the connection starts here and lands on the root node, not the other way round.

Key parameters

  • Required: the port (ai_embedding) has to be connected to a root node that turns text into vectors, and that root node will not run while the slot is empty.
Use cases
indexing a batch of documents into a store, then reading from that same store later in the workflow with the identical embedding node attached.
Parameters

Which setting does the node expose?

The Embeddings Oracle Database node has one parameter. For each one: the node as you configure it in n8n, what the parameter changes, and our field notes.

01

Model

model

What you see in n8n

Notes & use cases

Model is the only parameter on the node, and it decides which ONNX model produces the vectors. n8n calls it a resource locator, which means you either pick an entry from the list it loads for you or switch the selector to type an ID yourself. Use the second mode when the value comes from an earlier node.

Key parameters

  • Model: required, so leaving it empty stops the node before it runs. Choose from the list, or specify an ID, for instance {{ $json.model }} when a previous node carries the value.
Use cases
keeping one model ID in a config node and pointing both the indexing branch and the search branch at it, so the two never drift apart.
Need help

Need help automating Embeddings Oracle Database with n8n?

A person reads every message.

FAQ

Questions people ask next

01Is the Embeddings Oracle Database node included in n8n?
Yes. It ships inside the n8n AI package, so there is nothing to install and nothing extra to pay on the n8n side. It behaves the same on n8n Cloud, the hosted offer run by n8n, and on a self-hosted instance installed through Docker or npm under the Community Edition and its Sustainable Use licence. A workflow built on one runs on the other. The only cost that can appear is on the model provider side, since a provider bills its own API calls according to its own terms. n8n adds nothing on top of that.
02What do you need for the n8n Embeddings Oracle Database node to work?
Two things. First a credential, created once from the Credentials menu and reusable across every workflow on the instance, so the connection details stay out of the canvas. Second a root node with a free Embeddings slot, because this is a sub-node: it has no main input and never runs alone. Attach it to a node that needs vectors, an agent, a chain or a vector store, and fill the Model field. Model is marked required, so an empty value blocks execution before anything is sent.
03What are the limits of this node?
The node is at version 1 and exposes a single parameter, Model. There is no batching option, no retry setting and no way to change how the text is cut up before it is embedded, since that belongs to the root node or to a text splitter sub-node. A workflow saved on an earlier version of the node can show fewer options than the panel described here. No public n8n template currently uses this node either, so there is no ready-made workflow to copy from before you wire your own.
04What has to be plugged in, and where do you start?
At minimum, a root node. Sub-nodes only exist to serve one, so start from the node that actually consumes vectors, usually a vector store or a chain that searches documents, and add the embedding node underneath it. Fill Model first, connect the Embeddings port second, then run the root node and read the execution log: the sub-node call appears nested under it. Once that loop works, the same node can be duplicated onto the query side of the workflow with the identical model value.
05n8n or Make for this kind of setup?
It depends on where the data has to live and how you want to pay. Make is hosted only and billed per operation, which is predictable for a steady workflow and less so for a job that embeds thousands of documents in one run. n8n can be self-hosted through Docker or npm, so the text and the vectors stay on infrastructure you control, and it can also run on n8n Cloud when you would rather not manage a server. The visual logic differs too, so the honest test is to rebuild one real workflow on each.
Hack'celeration Lab

Get our weekly integration tips.

No spam. Unsubscribe anytime.

Same platform

Other docs you might need on this stack.

Browse all integrations