Static & SSR
Astro, Next.js, Nuxt, SvelteKit, Hugo or plain HTML. Files served from the edge; server routes render there too.
preview per branchinstant rollbackcustom domains + TLS
SITES · SERVER_APPS · DATABASES · WORKERS
Push a repo.
Get the whole app
running.
Your Laravel, Rails or Node app, your static and server-rendered sites, deployed straight from Git. The database, cache and queue workers they need live right next to them. No servers to patch, nothing to configure first, and one bill for all of it.
start for $0 →
then $20/mo · cancel anytime
Not ready to add a card? Static sites and uptime monitoring are free, no card. Or check if a site is down right now.
Coming from Forge, Heroku, Vercel or Netlify? Point us at the repo and we’ll detect the framework and bring your build settings with you.
$git push origin main
→detected laravel 12 · php 8.4
✓build image41.3s
✓new version up beside production9.8s
✓migrations ran · new version checked1.6s
✓visitors moved to the new version0.4s
✓production updated · visitors moved back12.2s
✓queue workers × 2 restarted3.6s
storefront.on-dply.app0 requests dropped
THE_DASHBOARD
Every app gets a live map of itself
Visitors, the edge in front, your app, and everything attached to it, on one screen. It sleeps when nobody is around, wakes on the next request, and scales out when traffic comes.
storefrontacme/storefront@mainLIVE
REQUESTS PER DAYlast 30 days
2,840
LIVE REQUESTS● live
GET/products/sneakers20041 msIAD
POST/cart20063 msLHR
GET/20018 msSJC
WHAT_RUNS_HERE
Everything an app needs, in one project
Start with a marketing site. Add the API, the database and the workers when you need them — same repo, same dashboard, same invoice.
PHP, Rails & Node
Your framework, unchanged, in a container. It scales out on traffic and sleeps when idle, so the meter stops too.
Laravel, Rails, Expressscale to zerologs + live tail
Databases & cache
Managed Postgres, MySQL, MongoDB and Valkey beside your app, plus SQLite at the edge for lighter work.
daily backupspoint-in-time restorequery console
Queues & jobs
Always-on queue workers that autoscale on backlog, with the scheduler built in and failed jobs one click from retry.
autoscalingworker groupsfailed-job alerts
ON_THE_EDGE
See every request, and where in the world it was answered
Each app's Traffic tab, as it looks for an example store: requests, failures and speed in plain words, and the Cloudflare locations that served its visitors.
TRAFFIC · LAST 7 DAYS
Your app answered 14,810 requests, about 2,116 a day, and sent 1.9 GB. Responses took 84 ms on average. Today so far: 1,342 requests, 3 failed.
Look closer
Busiest day in the last 30: Sep 27, with 3,912 requests ~2,116 / day
3 requests failed today 1,342
Responses take 84 ms on average 91.4% cached
Pages load in 0.94 s for real visitors LCP p75
Watch requests as they arrive ● Live
SERVED FROM · LAST 24 HOURS
Answered from 14 Cloudflare locations, most from Ashburn.
IADAshburn3,120SJCSan Jose2,480ORDChicago1,920LHRLondon1,650FRAFrankfurt1,210DFWDallas980HOW_IT_WORKS
No pipeline to assemble
We read the repo and fill in the rest. Nothing to write before the first deploy — a dply.yaml is there if you want one later.
01
Connect Git
GitHub, GitLab or Bitbucket. Deploy a branch on every push, or pin a tag.
02
Confirm what we found
Framework, build command, runtime, and whether it needs a server — monorepo packages included. Override anything.
03
Attach what it needs
A database, a cache, a queue. Credentials land in the environment; you don’t copy connection strings around.
04
Ship
A timed, streaming build log and an HTTPS hostname before it finishes. Every later push does it again.
DAY_TWO
The parts you only notice later
Preview branches
Every branch gets its own URL, with review comments pinned to the page.
Instant rollback
Every deploy is kept. Promote an old one back in a click — no rebuild.
Access rules
Password-gate staging, allow-list an office, rate-limit a path.
Real request logs
Live tail, CSV export, and Core Web Vitals from actual visitors.
Restore to a minute
Point-in-time recovery on managed databases, not just nightly dumps.
Alerts that matter
Failed jobs, crashing workers and downtime reach Slack, email or PagerDuty.
PRICING
Three plans. Unlimited sites. Pay for what runs.
Try Pro free for 5 days. On top of the plan, apps, databases and Valkey bill by the second they run, and requests and bandwidth by the unit — nothing is throttled. See full pricing →
Starter
For side projects
$0 /mo
- Unlimited sites
- 1 seat
- 2 databases · 2 queues
- 1 build at a time
Pro
For real projects
$29 /mo
- Unlimited sites
- 3 seats
- 10 databases · 10 queues
- Autoscaling workers
Team
For your whole company
$49 /mo
- Unlimited sites
- 10 seats · $5/extra
- 50 databases · 50 queues
- Audit log