mem0ai/mem0 · error · ImportError

The 'langchain_community' library is required. Please instal

Error message

The 'langchain_community' library is required. Please install it using 'pip install langchain_community'.

What it means

Module-level ImportError raised when mem0.vector_stores.langchain is imported but langchain_community is not installed. The module hard-depends on langchain_community.vectorstores.VectorStore at import time, so merely importing the module without the package fails.

Source

Thrown at mem0/vector_stores/langchain.py:9

import logging
from typing import Dict, List, Optional

from pydantic import BaseModel

try:
    from langchain_community.vectorstores import VectorStore
except ImportError:
    raise ImportError(
        "The 'langchain_community' library is required. Please install it using 'pip install langchain_community'."
    )

from mem0.vector_stores.base import VectorStoreBase

logger = logging.getLogger(__name__)


class OutputData(BaseModel):
    id: Optional[str]  # memory id
    score: Optional[float]  # distance
    payload: Optional[Dict]  # metadata


# Methods that accept a pre-computed embedding and return (Document, float) pairs.
# Tried in order; first match wins. Not part of the base VectorStore contract,
# but exposed by several concrete implementations.
_SCORED_BY_VECTOR_METHODS = [

View on GitHub (pinned to 001c235229)

Solutions

pip install langchain_community (or add it to requirements.txt / the Docker image)
Use a dedicated extra if provided (e.g. mem0ai[lc] style extras) so the dependency is declared
If langchain was unintended, switch the provider back to faiss/qdrant/chroma etc.

Example fix

# before
# provider='langchain' without the package -> ImportError at import time

# after
# shell: pip install langchain_community
from mem0.vector_stores.langchain import LCVectorStore

Defensive patterns

Strategy: fallback

Validate before calling

try:
    import langchain_community  # noqa
    HAVE_LC = True
except ImportError:
    HAVE_LC = False

provider = 'langchain' if HAVE_LC else 'faiss'

Try / catch

try:
    from mem0.vector_stores.langchain import LCVectorStore
except ImportError as e:
    raise RuntimeError('Run: pip install langchain_community') from e

Prevention

Declare langchain_community in requirements when using the langchain provider
Smoke-test provider imports in CI for every configured vector store

When it happens

Trigger: Selecting vector_store provider 'langchain' in MemoryConfig, or importing mem0.vector_stores.langchain, in an environment where pip install langchain_community was never run.

Common situations: Fresh installs using only 'pip install mem0ai' (langchain is an optional extra); deploying to slim Docker images; adding the langchain provider to an existing app without updating requirements.

Related errors

AI-assisted analysis of mem0ai/mem0@001c235229 (2026-08-15). Data as JSON: /api/errors/7d680a209564cb55. Report an issue: GitHub.