An AI-powered website builder. Describe any idea and get a complete, mobile-responsive website with HTML, CSS, and JS — ready to iterate, export, or deploy.
- AI generation via Gemini 2.5 / 3 Flash and Groq Llama 3.3
- Conversational refinement with chat history
- Mobile-first generated sites (touch targets, viewport meta, responsive layouts enforced)
- Cloud project save (Vercel Postgres) with delete + history clear
- ZIP export (drag into Vercel/Netlify/anywhere static)
This project uses Auth.js v5 (Google OAuth) and Vercel Postgres.
- In the Vercel dashboard, go to your project → Storage → Create → Postgres
- Once created, copy the
POSTGRES_URLvalues into.env.local
- Google Cloud Console → APIs & Services → Credentials → Create OAuth 2.0 Client ID
- Application type: Web
- Authorized redirect URIs:
http://localhost:3000/api/auth/callback/googlehttps://<your-prod-domain>/api/auth/callback/google
- Copy the Client ID and Secret
Copy .env.example to .env.local and fill in:
AUTH_SECRET=<openssl rand -base64 32>
AUTH_GOOGLE_ID=<from Google Console>
AUTH_GOOGLE_SECRET=<from Google Console>
POSTGRES_URL=<from Vercel>
POSTGRES_URL_NON_POOLING=<from Vercel>
GEMINI_API_KEY=<optional for AI>
GROQ_API_KEY=<optional for AI>
npm run db:migratenpm run devSign in with Google at http://localhost:3000/studio.
- Push this repo to GitHub.
- Import the repo in Vercel (
vercel.com/new). It autodetects Next.js — accept defaults. - Add env vars from
.env.examplein Project Settings → Environment Variables. Make sure to also add the production Google OAuth redirect URI in the Google Cloud Console. - Push to
mainto auto-deploy.