Overview
Add Mem0 to any ModuleX agent or workflow. Persistent agent memory management against the Mem0 REST API (api.mem0.ai) — add, search, and retrieve long-term memories so AI agents can recall user context, preferences, and prior conversations across sessions.
Categories: AI & Machine Learning · Knowledge Base · Agentic · Auth: API Key · Actions: 3
Authentication
API Key Authentication
Authenticate using your Mem0 API key1
Step 1
Go to https://app.mem0.ai and sign up or log in
2
Step 2
Open the dashboard and navigate to the ‘API Keys’ section
3
Step 3
Create a new API key or copy your existing one
4
Step 4
Paste the API key below
Required Credentials
Available Actions
add_memories — Add memories to Mem0 for persistent storage and retrieval.
add_memories — Add memories to Mem0 for persistent storage and retrieval.
Parameters
string
required
User ID associated with the memory (e.g., “user_123”, “alice@example.com”)
array
required
Array of message objects with role and content (e.g., [{“role”: “user”, “content”: “Hello”}])
Response
search_memories — Search for memories in Mem0 using semantic search.
search_memories — Search for memories in Mem0 using semantic search.
get_memories — Retrieve memories from Mem0 by ID or filter criteria.
get_memories — Retrieve memories from Mem0 by ID or filter criteria.
Parameters
string
required
User ID to retrieve memories for (e.g., “user_123”, “alice@example.com”)
string
Specific memory ID to retrieve (e.g., “mem_abc123”)
string
Start date for filtering by created_at (e.g., “2024-01-15”)
string
End date for filtering by created_at (e.g., “2024-12-31”)
integer
Maximum number of results to return (e.g., 10, 50, 100) (Default:
10)integer
Page number to retrieve for paginated list results (Default:
1)Response
Limits & Quotas
- Rate limits and quotas depend on your Mem0 plan; see https://docs.mem0.ai for current tiers.
add_memoriesis processed asynchronously — the response returns astatus(initiallyPENDING) and anevent_idfor polling progress, not the stored memory records themselves.- Error model: non-2xx responses and timeouts are caught and
returned as
success=False+errorrather than raising. Plan for retries on the agent side based on the error string.