
A starter template for integrating Mixedbread Search into ecommerce applications using Next.js, Vercel, and Mixedbread.
Note: This is a demo store that is currently not publicly accessible. We will add public stores in the future. Feel free to explore the code and use this as a starter template to build your own AI-powered search application.
https://github.com/user-attachments/assets/e57e6d13-10b0-40f7-9b3b-29e7d04c0de1
git clone <your-repo-url>cd vercel-integration-ecommerce
npm install
Create a .env file in the root directory:
cp .env.example .env
Add your Mixedbread API key to the .env file:
MXBAI_API_KEY=your-api-key-hereMXBAI_STORE_ID=your-store-id
To get your API key and store ID, you have two options:
From Vercel Integration (Recommended if deploying to Vercel):
From Mixedbread Platform (For standalone use):
This template is pre-configured to work with the commerce dataset. If you're using your own dataset, you'll need to update the code to match your metadata structure and content URLs.
Key files to update:
lib/types.ts - Update the ProductMetadata interface to match your dataset's metadata fieldscomponents/product-grid.tsx - Update references to metadata fields (e.g., metadata.name, metadata.description) and image_url.urlapp/api/search/route.ts - Adjust search parameters if needed (e.g., top_k, score_threshold)Example metadata structure used in this template:
{name: string;description?: string;filename?: string;price?: number;category?: string;// ... other fields}
The template also expects search results to include an image_url.url field for product images.
Start the development server:
npm run dev
Open http://localhost:3000 in your browser to see the application.
MIT