SurrealDB Agent Memory
Managed memory and knowledge layer for agents on SurrealDB: facts with provenance and supersession, hybrid retrieval, and a built-in MCP server.
Tags
Description
Conversations, documents and structured data become entities, attributes, relationships, instructions and sources stored together, with a tri-temporal model (system, known, valid time) in which corrections supersede old facts while the old fact is kept. Each Context serves REST and MCP (Streamable HTTP) at https://<context-host>/mcp with a static Bearer API key bound to that Context (no OAuth); tools: remember, recall, context, reflect, forget, upload, inspect. Python and TypeScript SDKs plus adapters for LangChain, OpenAI Agents, Vercel AI SDK and CrewAI. Plans: Sandbox free (3M tokens, one-off), Lite $30/month (1M tokens), Standard $300 (10M), Plus $1,100 (40M), Enterprise custom. SurrealDB reports a 77.2% BEAM 1M score.
Agent access
Related
- Mem0: Memory layer for AI agents and apps: drop-in persistent memory that lets agents learn from past interactions.
- Zep: Unified context layer that combines business data, documents and conversations into governed context and memory for agents.
- Letta: Platform for stateful agents with memory, identity and the ability to learn over time (formerly MemGPT); terminal UI, App Server, desktop app and Agent SDK.
- Hindsight: Open-source agent memory system from Vectorize focused on agents that learn: retain, recall and reflect operations over memory banks, with an MCP server and a managed cloud.
Sources
- https://surrealdb.com/agent-memory
- https://surrealdb.com/docs/agent-memory/integrations/mcp-server/install
- https://surrealdb.com/blog/introducing-surrealdb-agent-memory
- https://surrealdb.com/llms.txt
Machine-readable
- JSON
- Markdown
- Source file on GitHub (edit via pull request)