Mem0

Mem0 is a memory layer for AI agents and applications. Language models forget everything between sessions, so developers use Mem0 to store what a user has said or done and retrieve the relevant facts later, which lets an assistant stay personalized and consistent across conversations and across different agents. Integration is through an SDK for Python and Node, with a quickstart that adds a conversation as memories and then searches them by user. The vendor also lists an agent harness and plugin options, and documents an MCP server so coding agents and chat clients can use the same memory. It offers a hosted platform and has an open source project. The company recently launched DolphinBench, a benchmark for how agents use memory on real tasks. The audience is developers and teams building chatbots, support agents, copilots and other agentic products who want persistent context without designing their own storage and retrieval layer. Pricing is published in US dollars per month. Hobby is free and includes unlimited end users, 10,000 add requests and 1,000 retrieval requests per month, one project and community support. Starter is $19 per month with 50,000 add requests, 5,000 retrieval requests and one project. Pro is $249 per month with 500,000 add requests, 50,000 retrieval requests, unlimited projects, private Slack support and advanced analytics. Enterprise pricing is custom.