n8n-nodes-qualifire-model

Qualifire OpenAI‑Compatible Model for n8n Agents. This node injects the required X-Qualifire-Api-Key header and uses the Qualifire OpenAI‑compatible base URL, so you can connect it directly to the Agent → Chat Model input.

Base URL used by default:

https://proxy.qualifire.ai/api/providers/openai

Install (no-publish local)

docker-compose.yml

services:
  n8n:
    image: n8nio/n8n:latest
    ports: ["5678:5678"]
    environment:
      - N8N_CUSTOM_EXTENSIONS=/extensions
    volumes:
      - ./extensions:/extensions

Place the built package here:

./extensions/n8n-nodes-qualifire-model/

Then restart the container.

Build locally

npm i
npm run build
# copy dist/ into ./extensions/n8n-nodes-qualifire-model/

If your n8n version shows peer dep warnings, install matching versions: npm i -D n8n-workflow@<your-n8n-version> n8n-core@<your-n8n-version>

Usage

1. Add Qualifire Model node.
2. Set credentials:
   • OpenAI API (for Qualifire Model)
   • Qualifire API
3. (Optional) Toggle Responses API Mode if you prefer /responses over /chat/completions.
4. Connect the node's Model output to Agent → Chat Model.
5. Open chat and talk. The Agent will call OpenAI through Qualifire.

Test the node

Click Execute node on the Qualifire Model node. The first output (main) returns a live call with a test prompt. If it fails, check keys and ensure the URL does not include /v1.

Notes

• If your Agent refuses to connect to the node's Model output, your n8n build may expect a different output key (e.g., ai_model instead of aiModel). In that case, change the outputs array in QualifireModel.node.ts to ['main', 'ai_model'], rebuild, and try again.
• You can also switch the node to read keys from env by enabling Expose Bearer from Env Instead of Credential.