Integration guide

Add OpenLux to your n8n workflows

n8n automates pipelines with a visual editor, and its OpenAI node accepts a custom base URL. Point it at OpenLux and any workflow can add summarisation, classification, extraction or image generation as one more step.

Create an API key Read the docs

Abstract orange glass sculpture illustrating the n8n integration with OpenLux — n8n + OpenLux — one key, one endpoint, one bill.

Set it up in four steps

Create an OpenLux key

Create the key in the console. One key covers text, image, audio and retrieval models, so a workflow can mix steps without extra credentials.

Add OpenAI credentials with a base URL

In n8n, create an OpenAI credential and enable the base URL override, pointing it at the OpenLux endpoint.

Drop the model into a node

Use the OpenAI node — or an HTTP Request node — and name the model as the catalogue does: gpt-6-astra, gemini-3.8-flash, deepseek-v4, gpt-image-2-c.

Chain models inside one workflow

Because every modality shares one endpoint and one bill, a single run can classify with a cheap model and write with a flagship.

Configuration

Copy these blocks, replace the placeholder key and you are done.

OpenAI credentials in n8n

Credential: OpenAI API
Base URL:  https://openlux.lol/v1   (override enabled)

Node: OpenAI -> Message a Model
Model: gpt-6-astra
Prompt: {{ $json.text }}

HTTP Request node alternative

POST https://openlux.lol/v1/chat/completions
Authorization: Bearer sk-openlux-...
Content-Type: application/json

{
  "model": "deepseek-v4",
  "messages": [{ "role": "user", "content": "{{ $json.text }}" }]
}

Models that fit n8n

GPT-6 Astra

OpenAI · $0.3677 / 1M tokens

Gemini 3.8 Flash

Google · $0.0551 / 1M tokens

DeepSeek V4

DeepSeek · $0.0613 / 1M tokens

GPT Image 2 C

OpenAI · $0.00882 / request

The full list is in the model catalogue and the use-case collections.

n8n + OpenLux questions

Can n8n use a proxy or gateway instead of OpenAI directly?

Yes. The OpenAI credential has an optional base URL field, which is all that is needed to route calls through OpenLux.

Can a workflow use several models?

Yes. Different nodes can name different models, so a cheap classifier can feed a flagship writer inside the same execution.

How is cost attributed per workflow?

Every call is metered on the OpenLux side and shown per model in the console. Give a workflow its own key if you want a per-workflow view.

Does image generation work in a workflow too?

Yes. gpt-image-2-c and the Gemini image tiers are on the same endpoint, so a workflow can generate visuals and copy in one run.

Connect n8n to OpenLux today

One key covers every model in this guide and the rest of the catalogue, metered in USD with free trial credit to start.

Create an API key See pricing