Local development, trusted HTTPS
No public exposure. No self-signed certificate warnings.
Across clouds. Across the room.
Your machines. Your agents.
One connected deployment platform.
A self-hosted platform to build and deploy apps, and control routing across your connected machines. Let your agents handle deployments while you monitor and manage everything from one console.
$
sudo bash -c "$(curl -fsSL meshploy.com/install.sh)" ⌘
# Install Meshploy on my serverpaste into Claude Code, Codex or your own agent YOUR MACHINES, WORKING TOGETHER
Build here. Run there.
Fetch your code from Git. Build on your own machine. Publish the image to your registry, then run it on a cloud node—all connected over your private mesh.
- 01Fetch source
- 02Build locally
- 03Push image
- 04Deploy in cloud
YOUR MACHINES, WORKING TOGETHER
Give local services a public address.
Your service can stay on its office machine. Point a domain at your gateway and route requests across the mesh to where that service already runs.
- 01Your domain
- 02Gateway route
- 03Private mesh
- 04Existing service
YOUR MACHINES, WORKING TOGETHER
More replicas. More places to run.
Increase an application’s replica count and let Kubernetes schedule it across eligible cloud workers. Your gateway keeps providing the application’s public entry point.
- 01Set replicas
- 02Schedule workers
- 03Run containers
- 04Serve through gateway
⑂ Git repository↗ app.example.com
YOUR PRIVATE WIREGUARD MESH Different machines. Shared capabilities.
◇
GatewayPublic cloud · 100.64.0.1
PUBLIC ENTRY + IMAGE REGISTRY
HTTPS / TCP routing▤ Built-in registry
▱
Your laptopYour desk · 100.64.0.2
BUILD NODE
⑂ Source → Build image
▥
Cloud VM 01Provider A · 100.64.0.3
KUBERNETES WORKER
◇ App replica 1
⌘
Office machineYour network · 100.64.0.4
MESH-ONLY NODE
▧ Existing internal tool
▥
Cloud VM 02Provider B · 100.64.0.5
KUBERNETES WORKER
Available capacity◇ App replica 2
Selected workflowLines show logical flow over the mesh
Apache-2.0Infrastructure you controlPrivate WireGuard meshConsole · CLI · MCP
Meshploy is an open-source, self-hosted application platform. It brings deployment, private networking, routing, and resource management together across your servers—with a web console, CLI, API, and MCP tools for AI agents.
Community is open source under Apache-2.0. Talk to us about Enterprise availability and your organisation’s requirements.
01 MEET YOUR INFRASTRUCTURE WHERE IT IS
You don’t have to
start from scratch.
Already running something useful? Discover services on your nodes and route traffic to them. Deploy new applications alongside what you already have.
Existing containersHost servicesNew deployments
02 PRIVATE ACCESS. TRUSTED HTTPS.
Trusted HTTPS.
Private to your mesh.
A real domain and a browser-trusted Let’s Encrypt certificate for your local service—accessible only over your private mesh.
Private Let’s Encrypt certificates require DNS delegation. How it works ↗
Publish when you need to
Public HTTPS for your apps and external webhooks.
Your internal dependencies
Keep services inside the cluster when they need no external endpoint.
03 THE SAME PLATFORM. YOUR WAY IN.
Use the console.
Or bring your own workflow.
Explore visually, work from your terminal, or connect your coding agent. Each interface works with the same Meshploy resources.
meshploy
# One application definition. Your infrastructure. $ meshploy apply -f compose.yml ✓ Project resources reconciled ✓ Application configuration applied ✓ Routes configured # Inspect rollout status and logs with the CLI $ meshploy --help ▊
Compose + x-meshploy · Declarative deployment
YOU
Deploy the inventory app in my team’s project. Keep Postgres private.
I’ll inspect the project, apply the application configuration, and check its deployment.
01 Inspect project access ✓
02 Apply application manifest ✓
03 Check deployment status ✓
Application runninginventory.example.com
Your agent. Your project permissions. Your infrastructure.
04 BUILD WITH YOUR AGENT
From generated code
to a running app.
Give your coding agent a deployment destination. Connect through MCP, grant access to a project, and let it work with Meshploy’s deployment and management tools.
- First-class agent identities
- Project and resource permissions
- One declarative path through apply
YOU
Deploy the inventory app in my team’s project. Keep Postgres private.
I’ll inspect the project, apply the application configuration, and check its deployment.
01 Inspect project access ✓
02 Apply application manifest ✓
03 Check deployment status ✓
Application runninginventory.example.com
Your agent. Your project permissions. Your infrastructure.
05 AFTER THE FIRST DEPLOYMENT
Keep sight of
what you run.
Follow a deployment.
Inspect rollout history and logs when you need to understand a change.
Manage access.
Grant people and agents access to the projects and resources they need.
Plan for recovery.
Configure database backup schedules and work with saved snapshots.
Where Meshploy fits.
Start with where your apps run, how your machines connect, and what you want to operate yourself.
Vercel · Railway · Render
Hosted platformsDeployment model
Deploy applications on the provider’s platform, using its infrastructure and deployment workflows.
With Meshploy
Run on machines you own or rent, across providers and locations. You operate the hosts; Meshploy coordinates deployments and routing.
Coolify
Docker over SSHDeployment model
Manages Docker deployments on individual servers over SSH. For supported multi-server deployments, traffic distribution is configured separately.
With Meshploy
Makes connected machines work as one deployment cluster. Meshploy handles Kubernetes orchestration and gateway routing, so you manage apps without managing Kubernetes directly.
Dokploy
Traefik · Docker & SwarmDeployment model
Deploy apps with Docker and Swarm, then configure their domains and automatic HTTPS through Traefik.
With Meshploy
Manage routes and apps as independent resources. Connect routes to K3s workloads or existing services over the mesh, with Caddy handling on-demand TLS. Control traffic independently of deployment, without managing Kubernetes directly.
Uncloud
Decentralised · No central control planeDeployment model
Deploy Compose applications across a private mesh without Kubernetes or Swarm. Its decentralised design keeps healthy replicas running when a machine fails.
With Meshploy
Build, deploy, and manage apps through a console, CLI, or AI agent—without managing Kubernetes directly. Meshploy handles the underlying K3s orchestration, networking, and routing.
Currently uses a single gateway, which is a single point of failure. Gateway high availability is planned.
Is Meshploy a hosting provider?
Meshploy is software you install on infrastructure you own or rent. You provide and maintain the servers; Meshploy brings deployment, networking, and management together.
Do I need multiple servers?
You can start with one server and connect more as your needs grow. Different node roles let you choose how machines participate in your infrastructure.
Can an AI coding agent use Meshploy?
Yes. Meshploy provides MCP tools and agent identities. Connect a compatible client, give its agent the appropriate access, and use it to manage resources and deployments.
Can I connect applications I already run?
Yes. Discovery surfaces reported endpoints and containers on your nodes. Routes can point to supported existing services without first redeploying them in Meshploy.
What stays private?
You choose exposure through service access settings and routes. Mesh access keeps a service reachable on the private network; public routes expose the selected endpoint through the gateway.
Is Meshploy open source?
Meshploy Community is licensed under Apache-2.0. You can inspect the source, self-host it, and contribute on GitHub.
07 MAKE IT YOURS
Your next app.
Your infrastructure.
Bring a server. Create a project. Build something useful.
OPEN SOURCE
Built to be
yours.Apache-2.0 licensed
Self-hosted on your servers
Built to be
yours.Apache-2.0 licensed
Self-hosted on your servers