Now in Public Beta

Illuminate Your Data, Empower Your Generative AI

The transparent, high-performance RAG platform that lets you see exactly how your AI retrieves and reasons with your data. No more black boxes.

View on GitHub
$ npm install @brightrag/core

Trusted by developers at

VercelStripeLinearNotionFigma

The Bright Difference

RAG that you can actually understand

Built from the ground up for transparency, performance, and developer experience.

Transparent Retrieval
See exactly which documents were retrieved, their relevance scores, and why they matched. No more black-box AI.
High-Speed Indexing
Process millions of documents in seconds with our Rust-optimized vector indexing engine.
Developer-First API
Clean, intuitive SDK with TypeScript support. Get started with just 3 lines of code.
Hybrid Search
Combine vector similarity with keyword matching and metadata filters for precise retrieval.
Enterprise Security
SOC 2 compliant with encryption at rest and in transit. Your data stays yours.
Built-in Analytics
Track retrieval quality, latency metrics, and user satisfaction over time.

Interactive Sandbox

See RAG in action

Watch the entire retrieval process unfold in real-time. Every step is visible and explainable.

Retrieved Documents

Getting Started Guide95%

Learn how to set up BrightRAG in your project...

Vector Embeddings87%

Understanding how text is converted to vectors...

Hybrid Search82%

Combining semantic and keyword search for better results...

Retrieval Pipeline

Query Analysis
Vector Search
Re-ranking
Context Assembly

Generated Response

To implement RAG with BrightRAG, start by installing the SDK and initializing it with your API key. Then, index your documents using the index() method. Finally, use query() to retrieve relevant context and generate responses.

3 sources cited42ms latency

Integrations

Works with your stack

Connect to your favorite databases, LLMs, and frameworks with minimal configuration.

Vector Databases

PG
PostgreSQL
PC
Pinecone
WV
Weaviate
QD
Qdrant
MV
Milvus
RD
Redis

LLM Providers

OA
OpenAI
AN
Anthropic
CO
Cohere
MI
Mistral
OL
Ollama
GQ
Groq

Frameworks

NX
Next.js
EX
Express
FA
FastAPI
DJ
Django
FL
Flask
SP
Spring

Developer Experience

Ship faster with less code

Clean, intuitive APIs designed for developer productivity. Full TypeScript support included.

import { BrightRAG } from '@brightrag/core';

const rag = new BrightRAG({
  apiKey: process.env.BRIGHTRAG_API_KEY,
});

// Index your documents
await rag.index({
  documents: [
    { content: "Your document content...", metadata: { source: "docs" } }
  ]
});

// Query with full transparency
const result = await rag.query("How does RAG work?");

console.log(result.answer);
console.log(result.sources);     // Retrieved documents
console.log(result.scores);      // Relevance scores
console.log(result.reasoning);   // Why these docs were chosen

Ready to illuminate your AI?

Join thousands of developers building transparent, production-ready RAG applications. Get started for free today.

Star on GitHub