Open-source RAG infrastructure with SDKs and an API for developers to build accurate, cited AI chat and search on their own data.
What it does
Agentset is developer infrastructure for building RAG-powered AI chat and search. It handles extraction, chunking and retrieval so teams get accurate, cited answers without deep RAG expertise. It is multimodal (text, images, tables), model-agnostic (bring your own vector DB, embeddings and LLM), supports metadata filters and an MCP server, and ships JS/Python SDKs.
Core features
Managed RAG pipeline (extraction, chunking, retrieval)
Automatic source citations
Multimodal support (images, tables, graphs)
Model-agnostic: choose vector DB, embeddings, LLM
Metadata filtering
MCP server and AI SDK integration
22+ file-format ingestion
Best for
→Build a chatbot over private documents
→Add semantic or deep search to an app
→Ground answers in a large corpus with citations
→Ship production RAG without building it in-house
Pricing
Free
Pro
$49/mo
Toolspool rankingby monthly traffic
Reviews
Big-picture takes: what it's for and whether it delivers. High-engagement YouTube videos — not sponsored.