Quickstart
Go from zero to a working backend in under 5 minutes.
Prerequisites
You need a DYPAI account. Sign up free — no credit card required.
1. Create your project
Sign in and create an organization
Go to dypai.ai and sign in. Create an organization if you don't have one — it's the container for your projects.
Create a new project
Click New Project and choose a name. New projects start on the Free plan ($0); upgrade to Pro ($25/mo) anytime for custom domains, dev mode, and more apps. DYPAI provisions your database, auth, and storage automatically in seconds.
Or build in Studio (no IDE)
Don't want to wire up an IDE? Every project opens in Studio — the in-dashboard AI builder — by default. Describe your app in a chat, watch the live preview, and publish, all without leaving the dashboard.
2. Connect your IDE via MCP
The fastest way to build from your editor is by connecting your IDE to DYPAI using the Model Context Protocol (MCP). Your AI agent gets access to 40+ tools to create tables, endpoints, auth, and more.
Open the Conectar IA page
From your organization, open the Conectar IA (MCP tokens) page at /{org}/mcp/tokens and create a token.
Copy the configuration
Copy the JSON config for your client. Works with Claude Code, Codex, Cursor, Windsurf, Antigravity, and any MCP-compatible tool.
Paste into your IDE
Add the config to your IDE's MCP settings:
- Claude Code:
.mcp.jsonin your project root - Cursor:
.cursor/mcp.json - Codex:
.codex/config.toml
Start building
Ask your AI agent anything:
"Create a users table with email, name, and role"
"Build a POST /orders endpoint that saves to the database and sends an email"
"Set up authentication with Google OAuth"
AI Credits
Building and editing with AI spends AI Credits. On the Free plan you get +5 per day; Pro includes 100/month. Your deployed apps keep running regardless of your balance. Learn more.
CLI installer
Prefer the terminal? From your machine (with Node.js), run:
npx @dypai-ai/install
It walks you through choosing your IDE and writes the MCP config. For scripts or CI, use --client and --token — see npx @dypai-ai/install --help.
Download a starter template
Go to the Frontend tab in your project and download a starter template with the SDK and MCP pre-configured for your IDE.
3. Install the SDK
Connect your frontend to your DYPAI backend:
npm install @dypai-ai/client-sdk
Create the client:
// lib/dypai.ts
import { createClient } from '@dypai-ai/client-sdk'
export const dypai = createClient(
import.meta.env.VITE_DYPAI_URL // or process.env.NEXT_PUBLIC_DYPAI_URL
)
Your project URL is https://YOUR_PROJECT_ID.dypai.app. Find it in the project dashboard.
Security
Never expose API keys in browser code. The client SDK uses JWT authentication by default — no API key needed for user-facing apps.
4. Deploy your frontend New
Deploy your app to your-project.dypai.app. You have two easy paths:
- From your AI client — ask it to deploy. The local
@dypai-ai/mcppackage exposes adypai_deploy_productiontool that builds from your filesystem and ships to production. - From the dashboard — open the Frontend tab and click deploy.
Either way you get a live URL on a *.dypai.app subdomain (or your custom domain on Pro). Learn more about frontend hosting.