Mem0 (by Embedchain Inc.) provides a persistent memory infrastructure for AI agents and assistants via a simple API. It combines vector search, knowledge graph storage, and key-value caching to let agents remember user preferences, past interactions, and facts across sessions. Integration requires as few as three lines of code. In April 2026, Mem0 released a new token-efficient memory algorithm improving temporal query performance by 29.6 points. The company raised a $24M Series A to build the memory layer for AI. Key features: - Persistent memory API combining vector search, knowledge graph, and key-value storage - 3-line integration with any LLM framework or AI agent - Graph memory with multi-hop entity resolution (Pro tier) - OpenMemory: local-first MCP-compatible memory server for Claude Desktop, Cursor, Windsurf - SOC 2 Type I certified; HIPAA-compliant on Pro tier - Token-efficient memory algorithm with 29.6-point improvement on temporal queries (April 2026)
Hobby: free (10K memory adds, 1K retrievals/month). Starter: $19/month. Growth: $79/month. Pro: $249/month. Enterprise: custom.