mem0ai/mem0 · error · ImportError

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

Error message

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

What it means

Raised at import time when the `vecs` package (Supabase's pgvector client library) is not installed. mem0's Supabase vector store delegates all collection and query work to `vecs`, which is optional, so the module import fails fast with an install instruction.

Source

Thrown at mem0/vector_stores/supabase.py:10

import logging
import uuid
from typing import List, Optional

from pydantic import BaseModel

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

from mem0.configs.vector_stores.supabase import IndexMeasure, IndexMethod
from mem0.vector_stores.base import VectorStoreBase

logger = logging.getLogger(__name__)

VECS_MAX_QUERY_LIMIT = 1000


class OutputData(BaseModel):
    id: Optional[str]
    score: Optional[float]
    payload: Optional[dict]


class Supabase(VectorStoreBase):
    def __init__(
        self,

View on GitHub (pinned to 001c235229)

Solutions

Run `pip install vecs` in the environment running mem0.
Confirm the right package: it is `vecs` on PyPI (Supabase Vector), not `supabase`.
Ensure a Postgres/pgvector database URL is also configured, since vecs will need it at runtime (connection string with the pgvector extension available).

Example fix

# before
memory = Memory.from_config({"vector_store": {"provider": "supabase", ...}})  # ImportError

# after
# pip install vecs
memory = Memory.from_config({"vector_store": {"provider": "supabase", "config": {"collection_name": "mem", "connection_string": DB_URL}}})

Defensive patterns

Strategy: try-catch

Validate before calling

import importlib.util

if importlib.util.find_spec("vecs") is None:
    raise SystemExit("Missing dependency: vecs (not 'supabase'). Run: pip install vecs")

Try / catch

try:
    from mem0.vector_stores.supabase import Supabase
except ImportError as e:
    raise RuntimeError(f"Supabase vector client missing: {e}. Install with pip install vecs.") from e

Prevention

Remember the package is `vecs`, not `supabase` — document this in your setup README.
Pin vecs alongside mem0ai in requirements to keep pgvector protocol compatibility.
Ensure the Postgres connection string targets a database with the pgvector extension available.

When it happens

Trigger: `from mem0.vector_stores.supabase import Supabase` or Memory config with `"provider": "supabase"` where `import vecs` raises ImportError.

Common situations: Confusing the `supabase` JS/PY client (which is a different package) with `vecs`; base mem0ai installs; CI environments without the Supabase extra; new team members following the Supabase dashboard docs instead of mem0's setup docs.

Related errors

AI-assisted analysis of mem0ai/mem0@001c235229 (2026-08-15). Data as JSON: /api/errors/fd0b519c70bb23ef. Report an issue: GitHub.