Whitelisting IPs
IP whitelisting lets requests from registered source IPs skip Proxy-Authorization — easier for servers in fixed-egress environments. This page is the dashboard flow; for the API see Whitelists API.
Where the whitelist lives
The scope depends on the product:
Residential
Per sub-user
Mobile
Per sub-user
ISP
Per batch
A single source IP can be whitelisted in many places independently. There's no "global" whitelist.
Add an IP
Sub-user whitelist (gateway products)
Left nav → {Product} → Sub-users → {Sub-user} → Whitelisted IPs.
Add IP.
Enter:
IP or CIDR (max
/24for IPv4,/64for IPv6)Label (your description, e.g.
prod-eu-west-1)
Save.
Active within 60 seconds.
[screenshot: Whitelisted IPs panel with one entry]
ISP batch whitelist
Left nav → ISP → My IPs → {Batch} → Whitelisted IPs.
Add IP — same form.
The batch whitelist applies to every IP in that batch.
Remove
Click the trash icon next to an entry. Effective within 60 seconds.
CIDR vs single IP
Whitelist supports /24 (IPv4) or /64 (IPv6) at most.
One static server
Single 203.0.113.5
Container fleet behind NAT
The NAT IP (single entry)
Kubernetes egress
The NAT IP of the cluster's egress gateway
Multiple ASN-issued IPs from a single ISP
Several singles (or a /24 if contiguous)
Dynamic egress (laptops, autoscaling without NAT)
Don't use whitelist — use user/password auth
Limits
Per sub-user (per product)
50
Per ISP batch
50
A /24 counts as one entry; the gateway expands it server-side.
Finding your egress IP
To find what to whitelist, run from the source machine (without the proxy):
That's your source IP. Note that if your machine is behind a NAT or a load balancer, this returns the egress IP, which is what the gateway sees.
Mix whitelist and password
You can run both modes simultaneously per sub-user — whitelisted IPs skip auth, anyone else gets prompted for credentials. Useful when migrating a fleet to whitelisting in stages.
API alternative
Full schema: Whitelists API.
Common pitfalls
Whitelisted the wrong IP — when you
curl ipv4.icanhazip.comfrom your laptop you get the laptop's egress, not the production server's. SSH to the actual production machine first.CIDR too wide —
/16is rejected. Split into multiple/24entries or rethink whether you need that much.Forgot batch whitelist after a sub-user re-bind — moving an ISP batch to a new sub-user doesn't carry over the batch whitelist (it's per-batch, not per-sub-user, so it stays).
Last updated
Was this helpful?