v2.4.1 — now with ARM64 native builds
Ship without the
"works on my machine"
Isolated containers in milliseconds. Kubernetes-grade orchestration without the Kubernetes-grade headache. Your entire stack, green across every service.
1~/services $ deploy stack up --file deploy.yml --env production
2ℹ Reading deploy.yml — 12 service definitions found
3 [api-gateway] ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓ pulling image CACHE HIT
4 [auth-service] ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓ pulling image CACHE HIT
5 [worker-queue] ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓ pulling image CACHE HIT
6 [postgres-16] ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓ pulling image PULL 2.1s
7 [redis-cluster] ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓ pulling image CACHE HIT
8⟳ Routing network mesh — overlay: deploy-prod-net
9⟳ Injecting secrets from vault — 8 vars resolved
10⟳ Running health checks — waiting for readiness probes...
11✓ 12 services healthy · 47ms cold start · 0 failed probes
12✓ Stack live at https://prod.acme-corp.deploy.sh
13~/services $
scroll to compare
// platform comparison
⬆ Import Your Docker ComposeLet the data do the arguing
Toggle deployment scenarios. Watch metrics shift. No gate, no email wall — stress-test every claim before you sign up.
scenario:— 1 service, 512MB RAM
| Metric | Deploy this platform | Railwayrailway.app | Renderrender.com | Kubernetesself-managed |
|---|---|---|---|---|
Cold Start Time Time from deploy trigger to first healthy response | 47ms | 1.2s | 2.8s | 8–45s |
Memory Ceiling Maximum RAM per service instance | Unlimited | 8GB | 4GB | Node limit |
Concurrent Builds Parallel build pipelines running simultaneously | Unlimited | 2 | 1 | Manual config |
Rollback Speed Time to revert to last healthy deployment | < 1s | ~30s | ~60s | 2–5min |
Secret Injection How environment secrets are delivered to containers | Vault + env | Env vars only | Env vars only | K8s Secrets |
Zero-Downtime Deploy Blue/green or rolling deploy with no traffic interruption | Native | Partial* | Partial* | Manual config |
* Partial = requires manual configuration or has documented limitations. Benchmarks measured on equivalent instance sizes, December 2025.
all systems operational
Ready to watch your stack
go green?
No credit card. No cluster YAML. No 3 a.m. Slack alerts about a broken deploy pipeline. Just containers, spinning up in 47ms.
$ npm install -g @deploy/cli
$ deploy init
$ deploy up
✓ 47ms · stack live
✓ Free tier — no card required✓ SOC 2 Type II certified✓ 99.998% uptime SLA✓ GDPR compliant