A static IP for business is a fixed outbound address your servers and apps send traffic from. You need one when partners, ERPs, or payment providers allow-list IPs, or when server-to-server APIs authenticate by address instead of rotating tokens.
# one variable, fixed egress for every call
HTTPS_PROXY=https://user:pass@eu-01.outboundgateway.com:8443
# your ERP integration, as usual
curl -s https://erp.partner.example.com/v1/orders \
-H "Authorization: Bearer $TOKEN"
# the partner firewall sees the same two
# EU IPs on every request, every deploy
Most business internet connections ship with a dynamic address: the ISP assigns one from a pool and rotates it over time, and cloud platforms do the same to your servers on every redeploy. That is fine for browsing. It stops being fine the moment another system has to recognize you.
The comparison that matters for applications, not routers:
| Factor | Static IP | Dynamic IP |
|---|---|---|
| Control | You know the exact address your traffic leaves from and can hand it to anyone who asks. | The address belongs to the platform or ISP pool; you find out what it is after it changes. |
| Allow-listing | Partner firewalls, ERPs, and payment providers can pin your access to one known address. | Every rotation invalidates the allow-list entry; integrations break silently until someone updates the list. |
| Reliability | A fixed pair with failover keeps the same origin through redeploys and node maintenance. | Restarts, scale events, and lease renewals all reshuffle the address mid-transaction. |
| DNS and records | Stable egress keeps SPF-style records, reverse DNS plans, and monitoring targets valid long term. | Records tied to a dynamic address need constant re-pointing or wildcard tricks. |
| Cost | Cloud NAT routes run roughly 37 to 42 dollars a month at low volume in EU regions; a managed proxy runs from 19 euros. | Free with the connection, until the first outage or audit that a stable address would have prevented. |
A dynamic address is an ISP convenience. A static one is an operational commitment: it says your systems are recognizable, on purpose, by the people and firewalls that need to recognize them.
For the full background, read our blog post on static vs dynamic IPs.
The clearest sign you need a static IP address for business is an onboarding checklist that asks for one. Four situations produce that checklist.
ERPs and enterprise partners commonly open their API only to addresses you declare in advance. With a fixed pair, the integration keeps working through every deploy; with rotating egress, it breaks on a Friday night with nothing changed in your code.
Providers such as Adyen, GoCardless, and Mollie let you restrict API keys and webhooks to known source IPs, while payment gateways like CinetPay get one predictable, attributable origin for every API call. A fixed egress pair is how you use those restrictions without locking yourself out. See our Stripe static IP guide for a worked payment example.
Internal APIs, data platforms, and an increasing number of SaaS tools accept requests only from declared ranges. Pinned egress turns that policy into something your app satisfies by default, on any platform it runs on.
When a DPA or a customer questionnaire asks where data leaves the union, an egress address anchored in EU data centres answers it. The destination stays the destination; the origin becomes a fact you can state.
Running on a platform that rotates egress by default? We cover the specifics for Heroku and Supabase.
Three routes, with what each really costs once traffic and maintenance are counted.
| Option | Monthly cost | What you run | Watch out for |
|---|---|---|---|
| Cloud NAT gateway | ~$36.50 to $41.61 baseline in AWS (region-dependent), plus a per-GB processing fee and the public IPv4 charge | Nothing yourself; the cloud builds the path | Billed per gateway per AZ even at zero traffic; high availability doubles it |
| Self-hosted VM or VPN | ~$5 to $15 in hosting, plus your setup and patching time | A VM with a fixed address, or a VPN endpoint you maintain | One box to patch, monitor, and worry about; failover is your problem |
| OutboundGateway | From €19 (Starter, 10 GB, 20,000 requests) | Nothing; set HTTPS_PROXY and forget it |
HTTPS traffic only, which is what API integrations are |
Take a mid-size integration that moves 10 GB a month of API traffic, deployed on AWS Frankfurt. The NAT gateway route costs $0.052 per gateway-hour plus $0.005 per hour for the public IPv4 address, and $0.052 per GB processed. That is $37.96 for gateway hours, $3.65 for the address, and $0.52 for the data: $42.13 a month before any egress charges, and the bill is 99% fixed cost at that volume. Add a second gateway for availability and it doubles to $83.74.
The OutboundGateway Starter plan covers the same 10 GB for €19, about $21.66 at the October 2026 rate of €1 = $1.14, and includes a redundant pair of EU IPs with failover, TLS passthrough, and no per-gateway metering. Exchange rates move; the structure of the comparison does not.
What a fixed pair gives you either way
Honest limits: a proxy covers outbound HTTPS from your apps, not inbound traffic, not other protocols, and not the address your office router holds. Workloads in private subnets still need a route to the internet before they can reach any proxy.
Set one environment variable and every outbound call your systems make leaves from a fixed pair of EU addresses.
Starting from €19/month. Flexible plans for every scale. Cancel anytime.
Yes, the moment a partner, ERP, or payment provider asks you for an IP to allow-list, or a security review wants to know where your outbound traffic originates. If nothing on that list applies and you only browse and email, you can skip it. A proxy-based fixed IP pair starts at lower cost than the cloud alternatives and needs no infrastructure work.
A static IP stays the same address on every request and every restart. A dynamic IP changes when your platform recycles the connection, redeploys, or moves your workload. For allow-listing, audit trails, and DNS records that point at your egress, static is the only workable option.
Yes. The address your office router gets from the ISP is a separate concern from the address your servers and apps send from. What partner APIs see is your application egress, and you can pin that to a fixed pair of IPs with one HTTPS_PROXY environment variable, on any ISP or cloud platform.
Three typical routes: a cloud NAT gateway, roughly $37 to $42 a month at low volume in an EU region plus per-GB processing; a self-hosted VPN or VM at $5 to $15 a month in hosting plus your own setup and maintenance time; or a managed proxy like OutboundGateway from €19 a month with a redundant pair of EU IPs and no infrastructure to run.
Only when the other side cares about your address: partner APIs behind a firewall, ERPs, payment providers that restrict API keys to known IPs, or internal APIs that authenticate by origin. If every call is authenticated with tokens and the provider never looks at your IP, a static IP is optional, but it still makes logs and audits cleaner.
Happy to talk through how a fixed EU pair fits your stack.
Contact Our Founders →