list_memories
List Memories
Retrieve all memories with optional filtering by user, agent, app, or session. Supports pagination for large result sets. Use this to browse stored memories rather than searching by semantic similarity.
list_memories
Retrieve all memories with optional filtering by user, agent, app, or session. Supports pagination for large result sets. Use this to browse stored memories rather than searching by semantic similarity.
delete_memories
Delete one or more memories. You can delete a single memory by its ID, or bulk-delete all memories matching a scope filter (user, agent, app, or run). When using bulk delete, at least one scope filter is required.
get_memory
Retrieve a specific memory by its ID, optionally including its full change history. Use this to inspect a single memory's content, metadata, and version history.
delete_entity
Delete an entity and all of its associated memories. Supports user, agent, app, and run entity types. This permanently removes the entity and all memories scoped to it.
list_entities
List all entities (users, agents, apps, and runs/sessions) registered in Mem0. Optionally filter by entity type. Returns each entity's ID, name, type, memory count, and timestamps.
add_memory
Store new memories from conversation messages. Mem0 uses an LLM to extract facts from the messages and stores them as memories. You can optionally store messages directly without extraction by disabling inference. Memories can be scoped to a specific user, agent, app, or session (run). Supports graph memory for relationship extraction.
search_memories
Search memories using natural language queries with semantic similarity matching. Uses the v2 search API with support for advanced filtering using AND, OR, IN, comparison operators, and field-level filters. Returns ranked results based on relevance to the query.
update_memory
Update an existing memory's text content or metadata. Changes are versioned and tracked in the memory's history.
Store, retrieve, search, and manage persistent memories for AI applications across users, agents, and sessions. Add memories from conversations with automatic fact extraction or direct storage. Search memories using natural language with semantic similarity and advanced filtering. Manage graph-based memories with entity relationships and multi-hop reasoning. Organize memories with custom categories and instructions. Support multimodal content including images, PDFs, and markdown. Manage entities (users, agents, apps, runs), export memory data, and configure webhooks for memory event notifications.
Common questions about connecting Mem 0 to AI agents with Metorial.