Skip to main content
The Server stage is where all your configuration, document indexing, and preparation come together. In this guide, we’ll walk through the full app usage to create and launch your first RAG Server.

​
Full App Usage Workflow

Creating a RAG server with LarkupRAG is designed to be a seamless, end-to-end experience. By the time you reach the server generation stage, you have already:
  1. Configured your desired embedding models and vector database.
  2. Ingested your raw documents, websites, and files via ETL jobs.
  3. Indexed your clean data into a high-performance vector store.
Now, LarkupRAG takes your entire configuration and generates a minimal, deployable Node.js server.
Server Launch Interface

​
Generating the Server

Click Launch Server from the UI. The output directory contains everything you need, and absolutely nothing you don’t:
  • Zero Bloat: Only dependencies for your specific Vector Store are bundled.
  • No Build Step: Runs directly as a Node ESM backend (node server.mjs).
  • Ready to Ship: Includes a Dockerfile, docker-compose.yml, and vercel.json out of the box.
Generated Server Code

​
Watch the Walkthrough

To see the complete workflow in action—from configuring the vector store to launching the final standalone server—watch the demo video below:

​
Deployment Options

This lightweight backend can run anywhere:
Run the server container locally or deploy it to any cloud provider:
docker build -t my-rag-server .
docker run -p 8080:8080 my-rag-server
Push directly to Vercel using the included vercel.json:
cd .ragtoolkit/servers/<server-id>/gen
vercel --prod