Supermemory Documentation Index

Source: https://supermemory.ai/docs/

Core Concept

Supermemory functions as infrastructure for AI agent memory, enabling "perfect recall about users" to create more intelligent and personalized systems. The platform achieves state-of-the-art performance on benchmarks including LongMemEval and LoCoMo.

Key Components

• Agent memory systems
• Content extraction capabilities
• Connectors and data syncing
• Managed RAG platform

How It Operates

The system accepts diverse input formats -- text, files (PDF, images, documents), conversations, and video content. Supermemory then "intelligently indexes them and builds a semantic understanding graph" around entities like users or projects, retrieving contextually relevant information during queries.

Three Context Addition Methods

1. **Memory API** -- Extracts and evolves user facts in real-time, handling knowledge updates and temporal shifts to generate user profiles

2. **User Profiles** -- Combines static information (always-known facts) with dynamic, episodic details from recent conversations

3. **RAG Search** -- Advanced semantic retrieval featuring metadata filtering and contextual chunking

All three approaches utilize the same context pool when sharing a user ID, allowing flexible implementation strategies.

Documentation Sections

• Getting Started / Quickstart
• Authentication (API keys, scoped keys, connector branding)
• Introduction
• Content Management (documents, memories, search)
• Graph Memory (automatic memory evolution, knowledge updates, intelligent forgetting)
• User Profiles
• Connectors (GitHub, Gmail, Google Drive, Notion, OneDrive, S3, Web Crawler, Granola)
• Framework Integrations (LangChain, CrewAI, OpenAI SDK, Vercel AI SDK, LangGraph, 15+ more)
• SuperRAG (managed retrieval-augmented generation)
• SMFS (Semantic Memory File System)
• MCP (Model Context Protocol integration)
• MemoryBench (open-source benchmarking framework)
• Migration Guides (from Mem0 and Zep)
• API Reference (connections, container tags, content, documents, ingestion, profiles, search, settings)