Describe it.Watch it buildShip it live.

No. MicroDolphin is built for anyone who can describe what they want. If you can write a clear brief, you can build an app. That said, if you are a developer, you'll move even faster — the generated code is clean, readable, and ready for you to extend.

Absolutely. MicroDolphin handles the full stack for you. Describe what your app needs to do and our pipeline will scaffold the backend logic, set up the database schema, wire up the API, and deploy the whole thing. You don't need to know a line of Node, Python, or SQL to ship a complete, data-driven application.

Yes. You can specify your preferred language, framework, or tech stack in your brief and MicroDolphin will generate accordingly. Whether you want a Next.js frontend with a FastAPI backend, a full Django monolith, or a Go service — just say so. If you don't have a preference, we'll pick sensible defaults suited to your app's requirements.

Yes. If you need ongoing upkeep after launch, MicroDolphin offers optional auto-maintenance plans you can add to any project. These cover dependency updates, security patches, uptime monitoring, and periodic health checks — so your app stays current without you having to think about it.

When MicroDolphin generates your app, it deploys it to real Kubernetes infrastructure and gives you a unique URL — something like proj-abc123.preview.microdolphin.com. It's a real server running your real app. You can share that link with anyone, on any device, right now.

Most AI builders output code files. MicroDolphin outputs a running application — deployed and live before you've even finished reading the generated code. We also use a multi-agent pipeline instead of a single model, which means more reliable, architecturally coherent results on complex builds. And our RAG-powered context memory means your AI actually remembers your project across sessions.

The code MicroDolphin generates is clean, structured, and production-quality — not scaffolding or pseudo-code. Every generation goes through a quality-check agent before it's deployed. You can export it to GitHub, read every line, and continue building on it yourself. What you get is real code you own.

Your projects don't disappear. You can export all code to GitHub before or after cancellation. Any live previews on our infrastructure will be paused after the grace period, but your code is always yours to take.

Your code is stored securely and is only accessible to you (and your team, on Team plans). We don't train models on your code or share it with third parties. Enterprise plans include additional compliance controls for teams with stricter data requirements.

On the Team plan, yes — you can configure custom model preferences including API endpoints and system prompt overrides. This is designed for teams that have enterprise agreements with AI providers or want fine-grained control over model behavior.