Overview

Supermemory is the revolutionary memory and context engine for the AI era. It empowers AI applications with persistent, state-of-the-art memory capabilities, addressing the fundamental limitation of AI forgetting context between conversations. This highly scalable and performant system can be run entirely locally, offering unparalleled control and privacy. Supermemory is engineered to be the "Memory API for AI," providing a unified solution for context management, knowledge retrieval, and user personalization.

Key Features

• State-of-the-Art Memory: Excels in long-term memory across sessions, knowledge updates, and temporal reasoning, ranking #1 on major benchmarks like LongMemEval, LoCoMo, and ConvoMem.
• Intelligent Context Extraction: Automatically learns from conversations, extracting facts, resolving contradictions, and forgetting expired information to maintain a lean and relevant knowledge base.
• Automated User Profiles: Dynamically builds and maintains stable and recent user context, accessible in approximately 50ms.
• Hybrid Search: Seamlessly integrates Retrieval Augmented Generation (RAG) with personalized memory recall in a single query, delivering richer, more relevant results.
• Extensive Connectors: Effortlessly syncs data from Google Drive, Gmail, Notion, OneDrive, GitHub, and more, with real-time webhook support.
• Multi-modal Extraction: Processes PDFs, images (OCR), videos (transcription), and code (AST-aware chunking) to make diverse data types searchable.
• Local-First Deployment: Offers a zero-configuration, single-binary solution for self-hosting, with full offline capabilities using local models like Ollama.
• Unified API: Provides a singular, powerful API for developers to integrate memory, RAG, user profiles, connectors, and file processing into their AI products.
• Extensive Integrations: Features drop-in wrappers for popular AI frameworks including Vercel AI SDK, LangChain, and more.

Typical Use Cases

• Enhancing AI Tools: Integrate persistent memory into AI assistants like Claude Code, Cursor, and others, enabling them to remember user preferences, project details, and past discussions across conversations.
• Building Smarter AI Products: Developers can leverage the Supermemory API to embed robust memory, RAG, user profiling, and data connectors into their own AI agents and applications, streamlining development and eliminating complex infrastructure setup.
• Local and Private AI Deployments: Run the complete Supermemory engine on your own machine for maximum data privacy and control, compatible with various AI models, including fully offline options.
• Advanced Knowledge Management: Utilize Supermemory as the core context layer for sophisticated AI systems that require deep understanding and recall of user-specific information and external data sources.