ProductEcho is an AI-native runtime: your coding agent launches and manages your apps, databases and sites over the Model Context Protocol (MCP), from a first prototype to production scale. Connect Claude Code, Cursor or Codex once, and one prompt can launch a PostgreSQL database, deploy your code and return a public HTTPS URL.
Connect the ProductEcho MCP server to your coding agent. From the same chat it can launch PostgreSQL, deploy the app, inject the connection string and share a live URL.
Node.js, Next.js, Python, Java, Go and PHP apps, and static sites built with Vite, React, Vue or Svelte. Ruby, Rust and other languages run from a Dockerfile. For data there are managed PostgreSQL, MySQL and MongoDB databases, and WordPress hosting covers client sites. Your coding agent deploys apps over MCP and gets back a public HTTPS URL.
Yes. Free is $0 forever: 10 projects (dynamic apps and static sites), 10 custom domains with Auto SSL, a 1 GB PostgreSQL, MySQL or MongoDB database, and unlimited MCP keys and team members. No credit card. Upgrade to Pro when you need included infra credit, more projects, or metered compute.
Not on the default path. Docker packages an app and Kubernetes runs it; ProductEcho does both for you, so your coding agent can get you a first URL without you writing either. Languages without a built-in builder, such as Ruby and Rust, run from a Dockerfile you provide.
Yes. Connect a repository and deploy from that codebase. GitHub is a source, not a replacement for MCP.
Yes. ProductEcho provides custom domains, PostgreSQL, isolated runtimes, and role-based access (RBAC) on every plan. Pro adds included credit and metered compute. Enterprise includes a dedicated VPC (your own isolated network) and 24/7 dedicated support. Our uptime target is 99.9%, as set out in the Terms.