Porter is a platform that deploys and scales apps on Kubernetes inside your own AWS, GCP, or Azure account. CloudPloy deploys to a server you own or rent at cost, on a flat plan, and lets your AI agent deploy over MCP.
| Topic | Porter | CloudPloy |
|---|---|---|
| What it is | A platform that deploys and scales apps on Kubernetes inside your own AWS, GCP, or Azure account. | A hosted deploy control plane. Deploy from git or from your AI agent to servers you own or rent at cost |
| Pricing | $6 per GB RAM and $13 per vCPU per month, plus your cloud costs | Free plan, then $9.99, $19, or $39 per month. Compute billed separately at the provider rate |
| Where apps run | Kubernetes in your own AWS, GCP, or Azure account | Any server you own over SSH, or one you take from CloudPloy at the provider rate |
| Frameworks | Any framework, from Git or Docker | Node, Python, Go, Rust, Ruby, Java, PHP, and static sites via Nixpacks, plus Laravel and WordPress presets |
| Managed or self-hosted | Hosted control plane running workloads in your cloud | Hosted control plane; nothing to install or maintain on your side beyond your servers |
| Scaling | Autoscaling on usage or custom metrics | Up to 10 replicas of an app on its server; load balancing across servers is coming soon |
| Databases | Managed Postgres and Redis on AWS | Run databases such as MySQL on your server next to your apps |
| AI agent access | Remote MCP server and Porter CLI | HTTP MCP server on every plan, with OAuth sign-in and scoped tokens |
| Free tier | No free plan listed; startup program available | Free forever: 1 server, 1 app, up to 2 vCPU and 2 GB RAM |
Porter details checked in October 2026 on porter.run/pricing, docs.porter.run/mcp/overview, docs.porter.run/addons/datastores. They change; verify on Porter's site. CloudPloy numbers match our pricing page.
Porter is a platform that deploys and scales apps on Kubernetes inside your own AWS, GCP, or Azure account. CloudPloy is a hosted control plane: you connect a server you own, or take one at the provider rate, and deploy from git or from Claude Code, Cursor, or any MCP client.
Best fit for Porter: Teams that want Heroku-style deploys on Kubernetes in their own cloud, often for compliance. Best fit for CloudPloy: teams that want a flat plan, no markup on compute, and an AI agent that can deploy for them.
Connect your git repo to a CloudPloy server, copy your environment variables, deploy, then move your database and switch DNS.
Waitlist
Join 70+ enthusiasts on our waitlist!
Already in? Log in at app.cloudploy.com
