Install n8n first
Use n8n Cloud or run it locally. For a local test, install a Node.js version supported by n8n, start n8n with npx, open http://localhost:5678, then create the HTTP Request node shown below.
| What to install | Official source | Quick check |
|---|---|---|
| Node.js and npm | Node.js download | n8n npm installs require Node.js 20.19 through 24.x. Check with node --version and npm --version. |
| n8n local install | n8n npm install docs | npx n8n starts a local test instance at http://localhost:5678. |
| CorvusLLM key | Buy or top up balance | Use it in the Authorization header. |
node --version
npm --version
npx n8n
# Then open http://localhost:5678Zero-to-working setup
Use this block when you want the shortest reliable path from a clean install to one successful n8n request through CorvusLLM.
- n8n Cloud or local n8n
- Node.js version supported by n8n for local npm tests
- CorvusLLM key
Official source: n8n npm install docs, n8n HTTP Request node docs
Create an HTTP Request node and set Method, URL, Authorization, headers, and JSON body directly in the node.
- Add an HTTP Request node
- Set Method to POST and URL to https://base.corvusllm.com/v1/chat/completions
- Set Authorization Bearer header and JSON body
- Run the node once before adding loops or agents
- Inspect status code, response body, and error text in the execution view
The node execution returns HTTP 200 with a response body before loops, agents, or scheduled workflows are added.
| Value | Paste this |
|---|---|
| Method | POST |
| URL | https://base.corvusllm.com/v1/chat/completions |
| Authorization | Bearer YOUR_CORVUSLLM_KEY |
| Body model | gpt-5.5 |
Prefer HTTP Request node because it exposes URL, headers, and body directly.
HTTP Request node
Inside the workflow editor, add an HTTP Request node and set the request fields directly. This avoids hidden defaults from provider-specific nodes.
| n8n field | Where to change it | Value |
|---|---|---|
| Method | HTTP Request node > Parameters | POST |
| URL | HTTP Request node > URL | https://base.corvusllm.com/v1/chat/completions |
| Authentication | HTTP Request node > Authentication or Headers | Bearer token with your CorvusLLM key |
| Header | Headers table if you configure auth manually | Authorization: Bearer YOUR_CORVUSLLM_KEY |
| Body content type | Body settings | JSON |
{
"model": "gpt-5.5",
"messages": [
{
"role": "user",
"content": "Say n8n-ok"
}
]
}
Test and inspect response
Run the node once before placing it inside a larger workflow. In the n8n execution view, inspect the status code, response body, and any error text before adding loops, agents, or downstream processing.
| Result | Meaning | Next step |
|---|---|---|
| Text response appears | Auth, URL, model, and JSON body work. | Connect the node to the rest of the workflow. |
401 or auth error | Header or key is wrong. | Rebuild the Authorization header as Bearer YOUR_CORVUSLLM_KEY. |
429 | The key is out of balance. | Top up, then rerun the same node. |
Why this path
This path is explicit and reliable. You always know which URL, headers, and payload n8n is sending.