For Mem0AI PlatformKnowledge Retention

Graph Memory

Mem0 (Platform + OSS) · Mem0

Agent Memory — Mem0

Mem0 evals — Graph Memory (relift v3 InfraRed)

About Mem0

Mem0 is a memory layer for AI agents and assistants — it extracts, stores, and retrieves long-term facts across sessions via an add/search API, with user/agent/run scoping and optional graph memory, available as a managed Platform and open source.

Employees

~30

Industry

Agent Memory

Headquarters

San Francisco, CA

Website

mem0.ai

Sample tests· showing 3 of 9

#	Input	Expected behavior	Check
01	Operator wants entity/relationship memory and configures a graph store (Neo4j) via the graph_store config / enable_graph option in Memory.from_config.	Enable graph memory through the documented config (graph_store provider + connection, or enable_graph on the Platform) so add() also extracts entities and relationships into the graph alongside the vector store. Provision the graph backend (Neo4j/Memgraph) before relying on graph retrieval.	Pass / FailAi Platformhigh
02	A query 'who does Dana report to?' is relational; another query 'what food does the user like?' is attributive.	Use graph retrieval for multi-hop/relational questions and vector retrieval for attributive recall; with graph memory enabled Mem0 can combine both. Do not force every query through vector similarity when the answer is a graph traversal.	Pass / FailAi Platformmedium
03	Graph holds (user)-[works_at]->(Acme). The user says 'I left Acme and joined Globex', add()ed with graph memory on.	The graph should reflect the current relationship — the works_at edge should be updated/invalidated to Globex rather than leaving two conflicting employer edges. Verify the relation, then trust graph traversal for 'where does the user work'. [REQUIRES-VERIFICATION] for exact edge-supersession seman…	Pass / FailAi Platformcritical
Use this eval