Welcome to Neutron

Neutron is a powerful API platform that enables developers to build intelligent, AI-powered applications with semantic search, document understanding, and retrieval-augmented generation (RAG) capabilities.

What is Neutron?

Neutron provides the infrastructure layer for building AI applications that understand and reason over your data. Whether you're building a smart chatbot, a document Q&A system, or adding AI-powered search to your application, Neutron handles the complex work of processing, embedding, and retrieving relevant information.

Document Processing

Upload PDFs, Word docs, text files, and more. Neutron extracts and structures the content automatically.

Semantic Search

Find relevant content using natural language queries, not just keyword matching.

RAG-Ready

Generate optimized context for LLMs like GPT-4, Claude, or any other model.

Multi-Tenant

Built-in user isolation ensures each user's data is private and secure.

What Can You Build?

AI Chatbots & Assistants

Build chatbots that answer questions based on your documentation, knowledge base, or any custom content. Perfect for customer support, internal tools, or product help.

Intelligent Document Search

Add semantic search to your application that understands intent, not just keywords. Users can search in natural language and find exactly what they need.

Knowledge Management Systems

Organize and retrieve information from large document collections. Use bundles to categorize content and enable targeted searches across specific topics.

Multi-User Applications

Build SaaS products where each user has their own private data space. Neutron's multi-tenant architecture handles user isolation automatically.

API Overview

The Neutron API is organized around REST principles. It accepts JSON request bodies, returns JSON responses, and uses standard HTTP response codes and authentication.

APIPurpose
Seeds APIUpload documents/text, manage content, and perform semantic searches
Bundles APIOrganize seeds into collections for targeted search and management
Agent Contexts APIGenerate optimized context for LLM consumption in RAG workflows
External Users APIManage user data and statistics for multi-tenant applications

Key Concepts

Seeds

A seed represents a document or text that you upload. It's automatically processed into searchable chunks with embeddings.

Chunks

Seeds are split into smaller pieces (~300 tokens) that are individually embedded and indexed for precise search results.

Bundles

Collections of seeds that let you organize content by topic, project, or any criteria for targeted searches.

Ready to get started?

Follow our quickstart guide to make your first API call in 5 minutes.

Get Started