Skip to main content

memory.mem0_backend

Mem0 memory backend integration.

Provides long-term memory for agents using the Mem0 framework. Mem0 supports vector search, auto-deduplication, and multi-user isolation.

Requires mem0ai package: pip install mem0ai

Mem0Backend Objects​

class Mem0Backend(BaseMemoryBackend)

Memory backend powered by Mem0.

Parameters​

user_id : str User identifier for memory isolation. config : dict | None Mem0 configuration (vector store, embedding model, etc.). If None, uses Mem0 defaults. agent_id : str | None Optional agent identifier for agent-specific memories.

add​

async def add(messages: list[dict],
              metadata: dict[str, Any] | None = None) -> None

Add messages to Mem0 memory.

async def search(query: str, limit: int = 10) -> list[dict[str, Any]]

Search Mem0 memory for relevant entries.

close​

async def close() -> None

Release Mem0 resources.