Getting Started with EasyRAG

EasyRAG is a Retrieval-Augmented Generation (RAG) as a Service platform that makes it easy to build AI-powered document search and question-answering systems.

---

What is EasyRAG?

EasyRAG handles the complex infrastructure for RAG applications:

• Document Processing: Upload PDFs, Word docs, spreadsheets, presentations, and media files
• Automatic Chunking: Intelligent text splitting with configurable chunk sizes
• Vector Indexing: Embeddings stored in Qdrant for fast semantic search
• Transcription: Automatic speech-to-text for audio and video files
• Multi-Tenant: Built-in customer and dataset isolation
• Secure: Frontend tokens for safe browser access

---

Core Concepts

Datasets

A dataset is a logical container for related documents. Think of it as a project or knowledge base.

Examples:

• customer-support – Help documentation
• legal-contracts – Contract database
• user-123-personal – Per-user document storage

Files

Documents uploaded to a dataset. Supported formats:

• Documents: PDF, DOCX, XLSX, PPTX, CSV
• Media: MP3, WAV, MP4 (auto-transcribed)

Each file is:

• Chunked into smaller pieces (default: 300 tokens)
• Converted to vector embeddings
• Indexed for semantic search
• Stored with custom metadata

Authentication

Two authentication methods:

1. API Keys (Backend)

   • Full access to your account
   • Generate in the dashboard
   • Use for server-side operations

2. Frontend Tokens (Browser/Mobile)

   • Short-lived (max 24 hours)
   • Scoped to a single dataset
   • Created via /v1/tokens/create

---

Quick Start

1. Get Your API Key

Sign up at easyrag.com and generate an API key from the dashboard.

2. Upload Your First Document

bash
curl -X POST https://api.easyrag.com/v1/files/upload \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -F "datasetId=my-first-dataset" \
  -F "file=@document.pdf"

3. Search Your Document

bash
curl -X POST https://api.easyrag.com/v1/search \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "datasetId": "my-first-dataset",
    "question": "What is this document about?"
  }'

4. Get an AI Answer

bash
curl -X POST https://api.easyrag.com/v1/query \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "datasetId": "my-first-dataset",
    "question": "Summarize the key points"
  }'

---

API Endpoints

---

Use Cases

1. Customer Support Chatbot

Upload your help documentation and let users ask questions:

javascript
// Upload docs (once)
await uploadFiles('support-docs', helpFiles);

// User asks question
const answer = await query('support-docs', 'How do I reset my password?');

2. Document Management System

Let users search their uploaded documents:

javascript
// User uploads
await uploadFiles(`user-${userId}`, userFiles);

// User searches
const results = await search(`user-${userId}`, searchQuery);

3. Custom RAG Pipeline

Use EasyRAG for retrieval, bring your own LLM:

javascript
// Get relevant context
const chunks = await search('my-dataset', question);

// Use your own LLM
const answer = await yourLLM.complete({
  context: chunks.map(c => c.pageContent).join('\n'),
  question
});

4. Multi-Language Support

Upload documents in any language (embeddings support 100+ languages):

javascript
await uploadFiles('multilingual-docs', [
  'manual-en.pdf',
  'manual-es.pdf',
  'manual-fr.pdf'
]);

// Search works across all languages
const results = await search('multilingual-docs', 'comment réinitialiser');

---

Next Steps

1. Authentication – Learn about API keys and frontend tokens
2. Uploading Files – Detailed upload documentation
3. Search – Semantic search API
4. Query – AI-powered answers

---

Pricing

Pay-As-You-Go

• File Upload: 1 credit per file
• Search/Query: 0.1 credit per request
• Transcription: 0.1 credit per minute

Free Tier

• 20 credits on signup
• About 20 file uploads or 200 queries

Credit Packs

Purchase credits in the dashboard:

• Small: 100 credits ($5)
• Medium: 500 credits ($20)
• Large: 1200 credits ($40)

---

Base URL

All API requests use:

https://api.easyrag.com/v1

---

Support

• Documentation: easyrag.com/docs
• Dashboard: easyrag.com
• Email: support@easyrag.com

---

Limits

• Document files: 100 MB max
• Media files: 2 GB max
• Frontend tokens: 24 hour max TTL
• API rate limits: Contact support for enterprise limits

---

Technology Stack

EasyRAG is built on:

• Embeddings: OpenAI text-embedding-3-small
• LLM: GPT-4.1-mini
• Vector DB: Qdrant
• Storage: Google Cloud Storage
• Transcription: Whisper API
• Hosting: Firebase Cloud Functions + Cloud Run