n8n proxy setup
In n8n the HTTP Request node has its own Proxy option, so each node can leave from a different country or IP. Self-hosted n8n can also send all its traffic through a proxy.
- Protocols
- HTTP proxy on port 7000 with a login. n8n does not support SOCKS5
- Authentication
- Login and password inside the proxy URL
- Where the setting is
- HTTP Request node: Add option, then Proxy
- Works with
- Residential and mobile proxies. Written for n8n 2.47 (November 2026)
Before you start
- An n8n instance, cloud or self-hosted. The node option works on both; the environment variables below are for self-hosted n8n.
- Your proxy login and password from the Residential or Mobile page of the dashboard.
- Traffic on the product you use: a login without traffic gets 402.
- Host
gate.aeonproxy.com- Port (HTTP / HTTPS)
7000- Login
- your proxy login, with targeting keys if you need them:
LOGIN-cc-us-sid-abc123 - Password
- your proxy password
Copy the login and password from the Residential or Mobile page of the dashboard; the generator there also writes ready lines. Key syntax: targeting and sessions.
Proxy in the HTTP Request node
- Open the HTTP Request node and set the method and URL as usual.
- Under Options, press Add option and pick Proxy.
- Type the proxy URL with your login, as below.
- Run the node. To check the exit, send one GET to
https://api.ipify.org?format=json: theipin the answer is the proxy’s, not your server’s.
http://LOGIN-cc-us-sid-ab12cd34-time-1800:[email protected]:7000The node’s Proxy takes precedence over the instance’s proxy variables. Write http://: https:// would mean a TLS connection to the proxy itself, which port 7000 does not speak. A password with @ or : must be percent-encoded in the URL (%40, %3A).
One IP per workflow run
A rotating login gives every request a new IP. A run that signs in and then reads pages needs one IP from start to end, so give each run its own sticky key. The Proxy field takes expressions, and the run’s ID makes a good key:
http://LOGIN-cc-us-sid-{{ $execution.id }}-time-1800:[email protected]:7000Every HTTP Request node of the run with this URL leaves from the same exit for up to 30 minutes, and the next run gets a new IP.
Proxy for the whole instance
Self-hosted n8n reads HTTP_PROXY, HTTPS_PROXY and ALL_PROXY, and NO_PROXY for host names that must stay direct. Every node that uses n8n’s shared HTTP layer then goes through the proxy:
HTTP_PROXY=http://LOGIN-cc-us:[email protected]:7000
HTTPS_PROXY=http://LOGIN-cc-us:[email protected]:7000
NO_PROXY=db.internal,api.internalWith these variables everything n8n sends uses your traffic, calls to AI model APIs and your own services included, unless NO_PROXY lists them. Put the host names n8n uses for its own services there. For a few workflows, the node option is usually the better choice.
SOCKS5
n8n accepts HTTP proxies only. A socks5:// URL in the Proxy option is skipped without an error: the request goes out directly, or through the instance’s proxy variables, and the site sees your server’s IP. Use port 7000.
If it does not work
| What you see | What to check |
|---|---|
| The answer shows your server’s IP | A socks5:// URL (skipped), a typo in the scheme, or the Proxy option set on another node. |
407 or an authentication error | The login, the password or a key. A misspelt targeting value is refused like a wrong password. |
402 | No traffic left on the product. |
403 with an X-Reason header | A target blocked by default for every account (limits). |
Questions
Yes. Each HTTP Request node has its own Proxy option: give each its own login and -cc- key.
Each request goes through an exit in the country you chose, so it takes longer than a direct one, most of all for countries far from the target site.