Skip to content
RAG API

RAG API

Upload documents, we parse/chunk/embed/index them, you search with natural language. Production-quality retrieval in minutes, not months.
Get Started API Reference

Why RAG API?

Section titled “Why RAG API?”

Build AI-powered features without managing the retrieval stack. One API call to upload, one to search.

16 File Types

PDF, DOCX, PPTX, XLSX, images, Markdown, HTML, CSV, JSON, and more. We handle the parsing.

Quality-First Search

Voyage AI embeddings + cross-encoder reranking. nDCG@10 = 0.94 on SciFact benchmark.

Simple Data Model

Knowledge Base → Documents → Chunks. All search scoped to a KB. Metadata filtering built in.

SDKs Included

TypeScript and Python SDKs with full type safety. Or use the REST API directly.

How It Works

Section titled “How It Works”
  1. Create a Knowledge Base — a logical container for your documents.
  2. Upload documents — we parse, chunk, embed, and index them automatically.
  3. Search — send a natural language query, get ranked results with scores and metadata.
  4. Feed results to your LLM — that’s RAG.

Ready to start? Follow the Quickstart to get your first search result in 5 minutes.