Cloud hosting that holds 99.99% uptime under real traffic
One-click installs for the stacks you already run
WordPress
Laravel
Node.js
Next.js
Django
Ghost
Three things that cut our median page load to 380ms
NVMe storage, an edge cache in 34 cities, HTTP/3 on by default. The same WordPress build that served in 1.4s on shared hosting serves in 380ms here.
Daily backups, 30 day history
Every site is snapshotted at 03:00 in its own timezone and kept for 30 days. A full restore is one click and finishes in under four minutes.
A firewall that stops 99.98% of attacks
A managed WAF filters every request at 34 edge locations, patches known CVEs the day they are published, and pages our on-call engineer when traffic shape changes.
Scale to 8 vCPU without downtime
Move a site from 2 to 8 vCPU inside the dashboard. The resize runs live, keeps open connections, and bills by the hour from the moment it applies.
One console for every site, domain and backup you run
Anchorage runs the servers, the patching and the caching so your team stays on the product. Support answers in under six minutes at any hour, from engineers who can read your logs.
Sites overview
Deploy in 4 minutes
12 sites healthy
2 need attention
Managed WordPress, 380ms median
Create a site
99.99%
Uptime across all four regions, measured over the last 12 months
380ms
Median cached WordPress response, sampled hourly
5m 48s
Median first reply from support across 41,200 tickets last quarter
34
Edge locations serving cached assets
Support answers in 6 minutes, not 6 hours
Every plan includes 24/7 chat with engineers who hold root on the machines your site runs on. Median first reply last quarter was 5 minutes 48 seconds across 41,200 tickets. Advanced Support adds a named engineer and a 15 minute response target.
Read the support SLA
Move 40 sites across with zero downtime, at no charge
Our migration team moves as many sites as you have at no charge, runs each one on a staging URL first, and cuts DNS over only after you sign off.
Block storage
Attach up to 16TB of NVMe volumes to any server and grow them while the site stays online. Every volume is replicated three times inside the region, and a snapshot of 500GB completes in about 90 seconds.
See block storage pricing
Eight integrations that write themselves into your pipeline
Connect a repository and the next push deploys in about 40 seconds. Everything below is a toggle in the console, not a support ticket.
GitHub
Push to a branch and Anchorage builds and deploys it in about 40 seconds.
GitLab
The same pipeline as GitHub, with a preview URL on every merge request.
Cloudflare
Point your zone at us once and we keep the DNS records in sync from then on.
Stripe
Billing webhooks land on a dedicated worker, so checkout never queues behind page traffic.
Slack
Deploy, uptime and backup events post to the channel you name.
Sentry
Every error carries the release SHA and the server that threw it.
New Relic
The agent ships preinstalled and switched off. One toggle turns it on.
Let’s Encrypt
Certificates issue in under 60 seconds and renew 30 days before expiry.
Plans start at $69 a month and renew at the same price
Every plan includes one free domain for the first year, free SSL on every site, and free migration of what you already run.
Starter
For one business site with steady traffic
$
69
/mo
2 vCPU, 4GB RAM
Free migration of up to 3 sites
7 day metrics retention
50GB NVMe storage
Start a 14 day trial
No card needed to start
Most popular
Agency
For agencies running 20 to 40 client sites
$
99
/mo
8 vCPU, 16GB RAM
Free migration, unlimited sites
Unlimited sites on one plan
100GB NVMe storage
Isolated container per site
5TB bandwidth, then $0.01 per GB
Start a 14 day trial
No card needed to start
Scale
For teams that need staging and dedicated hardware
$
199
/mo
16 vCPU, 32GB RAM
250GB NVMe storage
Staging and one-click cloning on every site
Dedicated hardware, add $140 a month
Start a 14 day trial
No card needed to start
What $99 a month buys that $4 shared hosting does not
Measured on the same WordPress build, moved between all three in March.
Median WordPress response
380ms
1.4s
Whatever you tune it to
Uptime commitment
99.99%, paid in service credits
None published
None
Who patches the OS
We do, every week
The host, on a schedule you cannot see
You do
Migration
Free, unlimited sites, staged first
Manual, you do it
You do it
Backups
Daily, 30 days, one-click restore
Weekly, usually a paid add-on
You configure it
Scaling
2 to 16 vCPU live, billed hourly
You move to a bigger plan
Manual resize and a reboot
Support response
Median 5m 48s, engineers, 24/7
Tiered queue, hours to days
Community forums
Renewal price
Identical to your first term
Commonly 2x to 4x
Identical
Leaving
Server image and database dump, self-serve
cPanel export
It is your server
SOC 2 Type II, four regions, one container per site
The three questions a procurement review opens with, answered before you have to ask them.
SOC 2 Type II, audited every year
Our report covers the control set you will be asked about. Request it under NDA and it goes out the same working day.
Your data stays in the region you pick
Choose Frankfurt, Ashburn, Singapore or Sao Paulo when the server is created. Backups and read replicas never leave it.
Encrypted at rest, isolated at runtime
AES-256 on every volume and snapshot, TLS 1.3 at every edge, and one container per site so a compromise cannot read across.
What three customers measured after moving
“We moved 34 client sites over one weekend. Median load went from 2.1s to 0.6s and our Core Web Vitals pass rate went from 41% to 96% inside the first month.”
Sunita Raghavan
Head of Engineering, Northbeam Studio, Austin
“Checkout used to fall over on launch days. On Anchorage we held 18,000 concurrent sessions through Black Friday with no queue and no manual scaling.”
Marcus Delacroix
CTO, Fieldnote Commerce, Portland
“Support found a PHP-FPM memory leak and shipped the fix in 22 minutes at 2am. The same ticket sat with our old host for nine days.”
Sofia Almeida
Founder, Latitude Bakehouse, Lisbon
Six questions people ask before they move
If yours is not here, support answers in under six minutes at any hour.
The same. $69, $99 or $199 a month, and the yearly option keeps its 20% discount on renewal. Two things vary: bandwidth past 5TB bills at $0.01 per GB, and dedicated hardware on the Scale plan is a flat $140 a month. Both show as line items before they charge.
Your live site stays exactly where it is. We copy it onto a staging URL on our side, you test it there, and DNS moves only after you sign off. If something surfaces after cutover, your old host still has the site for as long as that account is open, so rolling back is a DNS change rather than a rescue operation.
A fresh WordPress install takes about 4 minutes. A single site migration is usually finished inside 2 hours. A batch of 40 client sites runs 2 to 3 business days, most of which is you reviewing the staging copies.
You pick the region when the server is created and nothing leaves it, backups included. Volumes and snapshots are AES-256 encrypted, traffic is TLS 1.3, and each site runs in its own container. Our engineers need a signed access request tied to your ticket before they can shell into your machine, and that request appears in your audit log within a minute.
You pull a full server image and a database dump from the console at any hour, with no ticket and no notice period. Billing stops at the end of the month you are in, transfer codes for domains are self-serve, and we do not hold DNS. Twenty-nine accounts left last quarter and the median time from clicking export to having the archive was 11 minutes.
The SLA pays 10% of that month bill for every 0.1% below 99.99%, up to the full month. We file the credit off our own monitoring, so it lands on your invoice without you having to notice the outage first.