mem0

Mem0 Platform Integration

Skill Graph: This skill is part of the Mem0 skill graph:

• mem0 (this skill) -- Platform Client SDK + OSS (Python + TypeScript)
• mem0-vercel-ai-sdk -- Vercel AI SDK provider

Mem0 is a managed memory layer for AI applications. It stores, retrieves, and manages user memories via API — no infrastructure to deploy. For self-hosted usage, see the OSS section in the client references below.

Step 1: Install and authenticate

Python:

pip install mem0ai
export MEM0_API_KEY="m0-your-api-key"

TypeScript/JavaScript:

npm install mem0ai
export MEM0_API_KEY="m0-your-api-key"

Get an API key at: https://app.mem0.ai/dashboard/api-keys?utm_source=oss&utm_medium=mem0-plugin-skill

Don't have a MEM0_API_KEY? Sign up at https://app.mem0.ai and create one from the dashboard. Keys start with m0-.

Step 2: Initialize the client

Python:

from mem0 import MemoryClient
client = MemoryClient(api_key="m0-xxx")

TypeScript:

import MemoryClient from 'mem0ai';
const client = new MemoryClient({ apiKey: 'm0-xxx' });

For async Python, use AsyncMemoryClient.

Step 3: Core operations

Every Mem0 integration follows the same pattern: retrieve → generate → store.

Add memories

messages = [
    {"role": "user", "content": "I'm a vegetarian and allergic to nuts."},
    {"role": "assistant", "content": "Got it! I'll remember that."}
]
client.add(messages, user_id="alice")

Search memories

results = client.search("dietary preferences", filters={"user_id": "alice"})
for mem in results.get("results", []):
    print(mem["memory"])

Get all memories

all_memories = client.get_all(filters={"user_id": "alice"})

Update a memory

client.update("memory-uuid", text="Updated: vegetarian, nut allergy, prefers organic")

Delete a memory

client.delete("memory-uuid")
client.delete_all(user_id="alice")  # delete all for a user

Common integration pattern

from mem0 import MemoryClient
from openai import OpenAI

mem0 = MemoryClient()
openai = OpenAI()

def chat(user_input: str, user_id: str) -> str:
    # 1. Retrieve relevant memories
    memories = mem0.search(user_input, filters={"user_id": user_id})
    context = "\n".join([m["memory"] for m in memories.get("results", [])])

    # 2. Generate response with memory context
    response = openai.chat.completions.create(
        model="gpt-5-mini",
        messages=[
            {"role": "system", "content": f"User context:\n{context}"},
            {"role": "user", "content": user_input},
        ]
    )
    reply = response.choices[0].message.content

    # 3. Store interaction for future context
    mem0.add(
        [{"role": "user", "content": user_input}, {"role": "assistant", "content": reply}],
        user_id=user_id
    )
    return reply

Common edge cases

• Search returns empty: v3 processes add() asynchronously — returns an event ID immediately. Wait 2-3s before searching. Also verify user_id matches exactly (case-sensitive) and use filters={"user_id": "..."} syntax.
• AND filter with user_id + agent_id returns empty: Entities are stored separately. {"AND": [{"user_id": "alice"}, {"agent_id": "bot"}]} returns nothing. Use OR instead, or query each separately.
• Duplicate memories: Don't mix infer=True (default) and infer=False for the same data. infer=True extracts facts via LLM with dedup. infer=False stores raw — same text can be stored twice.
• Implicit null scoping: filters={"user_id": "alice"} only returns memories where agent_id, app_id, run_id are ALL null. Wrap in {"OR": [...]} to include memories with non-null scoping fields.
• Platform vs OSS imports: Platform: from mem0 import MemoryClient. OSS: from mem0 import Memory. Don't mix them — MemoryClient talks to api.mem0.ai, Memory runs locally.
• v3 defaults: top_k=20, threshold=0.1, rerank=False. Adjust as needed.

v3 API (Current)

Mem0 v3 uses single-pass extraction, entity linking, and multi-signal retrieval.

Key v3 changes from v2:

• Endpoints: POST /v3/memories/add/, POST /v3/memories/search/, POST /v3/memories/ (paginated list)
• Extraction: Single ADD-only pass — no more UPDATE/DELETE operations during extraction. Memories accumulate rather than consolidate.
• Entity linking: Replaces graph memory. Auto-extracted during add(), no config needed. Remove enable_graph and graph_store from any old config.
• Defaults: top_k=20, threshold=0.1, rerank=False
• Removed params: org_id, project_id, enable_graph — all removed from SDK
• TypeScript: Exclusively camelCase (userId, agentId, appId, topK)
• Add response: Async — returns event ID immediately, poll via GET /v1/event/{event_id}/

See the migration guide for details.

Live documentation search

For the latest docs beyond what's in the references, use the doc search tool:

python ${CLAUDE_SKILL_DIR}/scripts/mem0_doc_search.py --query "topic"
python ${CLAUDE_SKILL_DIR}/scripts/mem0_doc_search.py --page "/platform/features/graph-memory"
python ${CLAUDE_SKILL_DIR}/scripts/mem0_doc_search.py --index

No API key needed — searches docs.mem0.ai directly.

Client SDK References

Language-specific deep references (Platform + OSS):

Platform References

Load these on demand for deeper detail:

Related Mem0 Skills