Mem0

Mem0 is a managed memory service that uses an LLM to extract, deduplicate, and manage facts from natural language input. memio provides two adapters for Mem0.

Installation

pip install memio[mem0]

This installs the mem0ai SDK as a dependency.

Setup

1. Create an account at mem0.ai.
2. Generate an API key from the dashboard.
3. Either pass the key directly or set the environment variable that the mem0ai SDK reads (see Mem0's own docs for the env var name).

Supported stores

Store	Adapter class
FactStore	Mem0FactAdapter
GraphStore	Mem0GraphAdapter

FactStore -- Mem0FactAdapter

Initialization

from memio.providers.mem0 import Mem0FactAdapter

facts = Mem0FactAdapter(api_key="m0-...")

You can also pass a config dict for advanced Mem0 configuration:

facts = Mem0FactAdapter(config={"api_key": "m0-...", ...})

Usage

from memio import Memio
from memio.providers.mem0 import Mem0FactAdapter

client = Memio(facts=Mem0FactAdapter(api_key="m0-..."))

# Add a fact
fact = await client.facts.add(
    content="Alice likes espresso",
    user_id="alice",
)
print(fact.id, fact.content)

# Search facts
results = await client.facts.search(query="coffee", user_id="alice")
for f in results:
    print(f.content, f.score)

# Get a specific fact
fact = await client.facts.get(fact_id=fact.id)

# Update a fact
updated = await client.facts.update(fact_id=fact.id, content="Alice likes cortados")

# List all facts for a user
all_facts = await client.facts.get_all(user_id="alice")

# Delete a single fact
await client.facts.delete(fact_id=fact.id)

# Delete all facts for a user
await client.facts.delete_all(user_id="alice")

GraphStore -- Mem0GraphAdapter

Initialization

from memio.providers.mem0 import Mem0GraphAdapter

graph = Mem0GraphAdapter(api_key="m0-...")

Usage

from memio import Memio, Triple
from memio.providers.mem0 import Mem0GraphAdapter

client = Memio(graph=Mem0GraphAdapter(api_key="m0-..."))

# Add triples
await client.graph.add(
    triples=[
        Triple(subject="Alice", predicate="works_at", object="Acme Corp"),
        Triple(subject="Alice", predicate="knows", object="Bob"),
    ],
    user_id="alice",
)

# Get triples for an entity
result = await client.graph.get(entity="Alice", user_id="alice")
for t in result.triples:
    print(f"{t.subject} -> {t.predicate} -> {t.object}")

# Search the graph
result = await client.graph.search(query="where does Alice work", user_id="alice")

# Get all triples
result = await client.graph.get_all(user_id="alice")

# Delete all graph data for a user
await client.graph.delete_all(user_id="alice")

Known quirks

Content rephrasing

Mem0's LLM may rephrase the content you provide. The fact.content you get back from add or get may not be the exact string you sent in.

Automatic deduplication

If you add content that Mem0 considers a duplicate of an existing fact, add returns an empty result set. The adapter raises a ProviderError wrapping a ValueError in this case, because there is no new fact to return.

No individual graph deletes

Mem0GraphAdapter.delete() is not supported. Calling it raises a ProviderError wrapping NotImplementedError. To remove graph data, use delete_all(user_id=...) instead, which clears all triples for the given user.