The Memory Layer for AI Agents - Drop-in memory infrastructure for AI agents and apps. Context that persists. Built for production.
Python agentic-memoryagentic-memory-systemagentsaiai-agentschatgptgenaillmlong-term-memorymemorymemory-managementpythonragstate-management
Add support for pgvector (#1675)
Dev Khant6cc4a31e910f
diff --git a/docs/components/vectordb.mdx b/docs/components/vectordb.mdxindex 5d5fede3e..288887c61 100644--- a/docs/components/vectordb.mdx+++ b/docs/components/vectordb.mdx@@ -9,6 +9,7 @@ Mem0 includes built-in support for various popular databases. Memory can utilize<CardGroup><Card title="Qdrant" href="#qdrant"></Card><Card title="Chroma" href="#chroma"></Card>+ <Card title="pgvector" href="#pgvector"></Card></CardGroup>@@ -22,6 +23,7 @@ To use Qdrant you can do like this:import osfrom mem0 import Memory+os.environ["OPENAI_API_KEY"] = "sk-xx"config = {"vector_store": {@@ -48,6 +50,7 @@ To use ChromaDB you can do like this:import osfrom mem0 import Memory+os.environ["OPENAI_API_KEY"] = "sk-xx"config = {"vector_store": {@@ -63,6 +66,34 @@ m = Memory.from_config(config)m.add("Likes to play cricket on weekends", user_id="alice", metadata={"category": "hobbies"})```+## pgvector++[pgvector](https://github.com/pgvector/pgvector) is open-source vector similarity search for Postgres. After connecting with postgres run `CREATE EXTENSION IF NOT EXISTS vector;` to create the vector extension.++Here's how to use it:++```python+import os+from mem0 import Memory++os.environ["OPENAI_API_KEY"] = "sk-xx"++config = {+ "vector_store": {+ "provider": "pgvector",+ "config": {+ "user": "test",+ "password": "123",+ "host": "127.0.0.1",+ "port": "5432",+ }+ }+}++m = Memory.from_config(config)+m.add("Likes to play cricket on weekends", user_id="alice", metadata={"category": "hobbies"})+```+## Common issues### Using model with different dimensionsdiff --git a/mem0/configs/vector_stores/pgvector.py b/mem0/configs/vector_stores/pgvector.pynew file mode 100644index 000000000..41134dfe8--- /dev/null+++ b/mem0/configs/vector_stores/pgvector.py@@ -0,0 +1,34 @@+from typing import Optional, Dict, Any++from pydantic import BaseModel, Field, model_validator++class PGVectorConfig(BaseModel):++ dbname: str = Field("postgres", description="Default name for the database")+ collection_name: str = Field("mem0", description="Default name for the collection")+ embedding_model_dims: Optional[int] = Field(1536, description="Dimensions of the embedding model")+ user: Optional[str] = Field(None, description="Database user")+ password: Optional[str] = Field(None, description="Database password")+ host: Optional[str] = Field(None, description="Database host. Default is localhost")+ port: Optional[int] = Field(None, description="Database port. Default is 1536")++ @model_validator(mode="before")+ def check_auth_and_connection(cls, values):+ user, password = values.get("user"), values.get("password")+ host, port = values.get("host"), values.get("port")+ if not user and not password:+ raise ValueError("Both 'user' and 'password' must be provided.")+ if not host and not port:+ raise ValueError("Both 'host' and 'port' must be provided.")+ return values++ @model_validator(mode="before")+ @classmethod+ def validate_extra_fields(cls, values: Dict[str, Any]) -> Dict[str, Any]:+ allowed_fields = set(cls.model_fields.keys())+ input_fields = set(values.keys())+ extra_fields = input_fields - allowed_fields+ if extra_fields:+ raise ValueError(f"Extra fields not allowed: {', '.join(extra_fields)}. Please input only the following fields: {', '.join(allowed_fields)}")+ return values+\ No newline at end of filediff --git a/mem0/memory/main.py b/mem0/memory/main.pyindex 81bbb65f0..142f6b7a2 100644--- a/mem0/memory/main.py+++ b/mem0/memory/main.py@@ -94,7 +94,6 @@ class Memory(MemoryBase):])existing_memories = self.vector_store.search(- name=self.collection_name,query=embeddings,limit=5,filters=filters,@@ -169,7 +168,7 @@ class Memory(MemoryBase):dict: Retrieved memory."""capture_event("mem0.get", self, {"memory_id": memory_id})- memory = self.vector_store.get(name=self.collection_name, vector_id=memory_id)+ memory = self.vector_store.get(vector_id=memory_id)if not memory:return None@@ -210,9 +209,7 @@ class Memory(MemoryBase):filters["run_id"] = run_idcapture_event("mem0.get_all", self, {"filters": len(filters), "limit": limit})- memories = self.vector_store.list(- name=self.collection_name, filters=filters, limit=limit- )+ memories = self.vector_store.list(filters=filters, limit=limit)excluded_keys = {"user_id", "agent_id", "run_id", "hash", "data", "created_at", "updated_at"}return [@@ -258,9 +255,7 @@ class Memory(MemoryBase):capture_event("mem0.search", self, {"filters": len(filters), "limit": limit})embeddings = self.embedding_model.embed(query)- memories = self.vector_store.search(- name=self.collection_name, query=embeddings, limit=limit, filters=filters- )+ memories = self.vector_store.search(query=embeddings, limit=limit, filters=filters)excluded_keys = {"user_id", "agent_id", "run_id", "hash", "data", "created_at", "updated_at"}@@ -330,7 +325,7 @@ class Memory(MemoryBase):)capture_event("mem0.delete_all", self, {"filters": len(filters)})- memories = self.vector_store.list(name=self.collection_name, filters=filters)[0]+ memories = self.vector_store.list(filters=filters)[0]for memory in memories:self._delete_memory_tool(memory.id)return {'message': 'Memories deleted successfully!'}@@ -358,7 +353,6 @@ class Memory(MemoryBase):metadata["created_at"] = datetime.now(pytz.timezone('US/Pacific')).isoformat()self.vector_store.insert(- name=self.collection_name,vectors=[embeddings],ids=[memory_id],payloads=[metadata],@@ -367,9 +361,7 @@ class Memory(MemoryBase):return memory_iddef _update_memory_tool(self, memory_id, data, metadata=None):- existing_memory = self.vector_store.get(- name=self.collection_name, vector_id=memory_id- )+ existing_memory = self.vector_store.get(vector_id=memory_id)prev_value = existing_memory.payload.get("data")new_metadata = metadata or {}@@ -387,7 +379,6 @@ class Memory(MemoryBase):embeddings = self.embedding_model.embed(data)self.vector_store.update(- name=self.collection_name,vector_id=memory_id,vector=embeddings,payload=new_metadata,@@ -397,18 +388,16 @@ class Memory(MemoryBase):def _delete_memory_tool(self, memory_id):logging.info(f"Deleting memory with {memory_id=}")- existing_memory = self.vector_store.get(- name=self.collection_name, vector_id=memory_id- )+ existing_memory = self.vector_store.get(vector_id=memory_id)prev_value = existing_memory.payload["data"]- self.vector_store.delete(name=self.collection_name, vector_id=memory_id)+ self.vector_store.delete(vector_id=memory_id)self.db.add_history(memory_id, prev_value, None, "DELETE", is_deleted=1)def reset(self):"""Reset the memory store."""- self.vector_store.delete_col(name=self.collection_name)+ self.vector_store.delete_col()self.db.reset()capture_event("mem0.reset", self)diff --git a/mem0/utils/factory.py b/mem0/utils/factory.pyindex 1569c1fa1..d3397567e 100644--- a/mem0/utils/factory.py+++ b/mem0/utils/factory.py@@ -52,6 +52,7 @@ class VectorStoreFactory:provider_to_class = {"qdrant": "mem0.vector_stores.qdrant.Qdrant","chroma": "mem0.vector_stores.chroma.ChromaDB",+ "pgvector": "mem0.vector_stores.pgvector.PGVector"}@classmethoddiff --git a/mem0/vector_stores/chroma.py b/mem0/vector_stores/chroma.pyindex 1399f08b8..6a2402b82 100644--- a/mem0/vector_stores/chroma.py+++ b/mem0/vector_stores/chroma.py@@ -54,6 +54,7 @@ class ChromaDB(VectorStoreBase):self.client = chromadb.Client(self.settings)+ self.collection_name = collection_nameself.collection = self.create_col(collection_name)def _parse_output(self, data):@@ -109,12 +110,11 @@ class ChromaDB(VectorStoreBase):)return collection- def insert(self, name, vectors, payloads=None, ids=None):+ def insert(self, vectors, payloads=None, ids=None):"""Insert vectors into a collection.Args:- name (str): Name of the collection.vectors (list): List of vectors to insert.payloads (list, optional): List of payloads corresponding to vectors. Defaults to None.ids (list, optional): List of IDs corresponding to vectors. Defaults to None.@@ -122,12 +122,11 @@ class ChromaDB(VectorStoreBase):self.collection.add(ids=ids, embeddings=vectors, metadatas=payloads)- def search(self, name, query, limit=5, filters=None):+ def search(self, query, limit=5, filters=None):"""Search for similar vectors.Args:- name (str): Name of the collection.query (list): Query vector.limit (int, optional): Number of results to return. Defaults to 5.filters (dict, optional): Filters to apply to the search. Defaults to None.@@ -139,23 +138,21 @@ class ChromaDB(VectorStoreBase):final_results = self._parse_output(results)return final_results- def delete(self, name, vector_id):+ def delete(self, vector_id):"""Delete a vector by ID.Args:- name (str): Name of the collection.vector_id (int): ID of the vector to delete."""self.collection.delete(ids=vector_id)- def update(self, name, vector_id, vector=None, payload=None):+ def update(self, vector_id, vector=None, payload=None):"""Update a vector and its payload.Args:- name (str): Name of the collection.vector_id (int): ID of the vector to update.vector (list, optional): Updated vector. Defaults to None.payload (dict, optional): Updated payload. Defaults to None.@@ -163,12 +160,11 @@ class ChromaDB(VectorStoreBase):self.collection.update(ids=vector_id, embeddings=vector, metadatas=payload)- def get(self, name, vector_id):+ def get(self, vector_id):"""Retrieve a vector by ID.Args:- name (str): Name of the collection.vector_id (int): ID of the vector to retrieve.Returns:@@ -186,33 +182,24 @@ class ChromaDB(VectorStoreBase):"""return self.client.list_collections()- def delete_col(self, name):- """- Delete a collection.+ def delete_col(self):+ """ Delete a collection. """+ self.client.delete_collection(name=self.collection_name)- Args:- name (str): Name of the collection to delete.- """- self.client.delete_collection(name=name)-- def col_info(self, name):+ def col_info(self):"""Get information about a collection.- Args:- name (str): Name of the collection.-Returns:dict: Collection information."""- return self.client.get_collection(name=name)+ return self.client.get_collection(name=self.collection_name)- def list(self, name, filters=None, limit=100):+ def list(self, filters=None, limit=100):"""List all vectors in a collection.Args:- name (str): Name of the collection.filters (dict, optional): Filters to apply to the list.limit (int, optional): Number of vectors to return. Defaults to 100.diff --git a/mem0/vector_stores/configs.py b/mem0/vector_stores/configs.pyindex 28e4e1c40..6308afb39 100644--- a/mem0/vector_stores/configs.py+++ b/mem0/vector_stores/configs.py@@ -13,7 +13,8 @@ class VectorStoreConfig(BaseModel):_provider_configs: Dict[str, str] = {"qdrant": "QdrantConfig",- "chroma": "ChromaDbConfig"+ "chroma": "ChromaDbConfig",+ "pgvector": "PGVectorConfig"}@model_validator(mode="after")diff --git a/mem0/vector_stores/pgvector.py b/mem0/vector_stores/pgvector.pynew file mode 100644index 000000000..66ffa9268--- /dev/null+++ b/mem0/vector_stores/pgvector.py@@ -0,0 +1,241 @@+import json+from typing import Optional, List, Dict, Any+from pydantic import BaseModel++try:+ import psycopg2+ from psycopg2.extras import execute_values+except ImportError:+ raise ImportError("PGVector requires extra dependencies. Install with `pip install psycopg2`") from None+++from mem0.vector_stores.base import VectorStoreBase++class OutputData(BaseModel):+ id: Optional[str]+ score: Optional[float]+ payload: Optional[dict]+++class PGVector(VectorStoreBase):+ def __init__(+ self,+ dbname,+ collection_name,+ embedding_model_dims,+ user,+ password,+ host,+ port+ ):+ """+ Initialize the PGVector database.++ Args:+ dbname (str): Database name+ collection_name (str): Collection name+ embedding_model_dims (int): Dimension of the embedding vector+ user (str): Database user+ password (str): Database password+ host (str, optional): Database host+ port (int, optional): Database port+ """+ self.collection_name = collection_name++ self.conn = psycopg2.connect(+ dbname=dbname,+ user=user,+ password=password,+ host=host,+ port=port+ )+ self.cur = self.conn.cursor()++ collections = self.list_cols()+ if collection_name not in collections:+ self.create_col(embedding_model_dims)++ def create_col(self, embedding_model_dims):+ """+ Create a new collection (table in PostgreSQL).++ Args:+ name (str): Name of the collection.+ embedding_model_dims (int, optional): Dimension of the embedding vector.+ """+ self.cur.execute(f"""+ CREATE TABLE IF NOT EXISTS {self.collection_name} (+ id UUID PRIMARY KEY,+ vector vector({embedding_model_dims}),+ payload JSONB+ );+ """)+ self.conn.commit()++ def insert(self, vectors, payloads = None, ids = None):+ """+ Insert vectors into a collection.++ Args:+ vectors (List[List[float]]): List of vectors to insert.+ payloads (List[Dict], optional): List of payloads corresponding to vectors.+ ids (List[str], optional): List of IDs corresponding to vectors.+ """+ json_payloads = [json.dumps(payload) for payload in payloads]++ data = [(id, vector, payload) for id, vector, payload in zip(ids, vectors, json_payloads)]+ execute_values(self.cur, f"INSERT INTO {self.collection_name} (id, vector, payload) VALUES %s", data)+ self.conn.commit()++ def search(self, query, limit = 5, filters = None):+ """+ Search for similar vectors.++ Args:+ query (List[float]): Query vector.+ limit (int, optional): Number of results to return. Defaults to 5.+ filters (Dict, optional): Filters to apply to the search. Defaults to None.++ Returns:+ list: Search results.+ """+ filter_conditions = []+ filter_params = []++ if filters:+ for k, v in filters.items():+ filter_conditions.append(f"payload->>%s = %s")+ filter_params.extend([k, str(v)])++ filter_clause = "WHERE " + " AND ".join(filter_conditions) if filter_conditions else ""++ self.cur.execute(f"""+ SELECT id, vector <-> %s::vector AS distance, payload+ FROM {self.collection_name}+ {filter_clause}+ ORDER BY distance+ LIMIT %s+ """, (query, *filter_params, limit))++ results = self.cur.fetchall()+ return [OutputData(id=str(r[0]), score=float(r[1]), payload=r[2]) for r in results]++ def delete(self, vector_id):+ """+ Delete a vector by ID.++ Args:+ vector_id (str): ID of the vector to delete.+ """+ self.cur.execute(f"DELETE FROM {self.collection_name} WHERE id = %s", (vector_id,))+ self.conn.commit()++ def update(self, vector_id, vector = None, payload = None):+ """+ Update a vector and its payload.++ Args:+ vector_id (str): ID of the vector to update.+ vector (List[float], optional): Updated vector.+ payload (Dict, optional): Updated payload.+ """+ if vector:+ self.cur.execute(f"UPDATE {self.collection_name} SET vector = %s WHERE id = %s", (vector, vector_id))+ if payload:+ self.cur.execute(f"UPDATE {self.collection_name} SET payload = %s WHERE id = %s", (psycopg2.extras.Json(payload), vector_id))+ self.conn.commit()++ def get(self, vector_id) -> OutputData:+ """+ Retrieve a vector by ID.++ Args:+ vector_id (str): ID of the vector to retrieve.++ Returns:+ OutputData: Retrieved vector.+ """+ self.cur.execute(f"SELECT id, vector, payload FROM {self.collection_name} WHERE id = %s", (vector_id,))+ result = self.cur.fetchone()+ if not result:+ return None+ return OutputData(id=str(result[0]), score=None, payload=result[2])++ def list_cols(self) -> List[str]:+ """+ List all collections.++ Returns:+ List[str]: List of collection names.+ """+ self.cur.execute("SELECT table_name FROM information_schema.tables WHERE table_schema = 'public'")+ return [row[0] for row in self.cur.fetchall()]++ def delete_col(self):+ """ Delete a collection. """+ self.cur.execute(f"DROP TABLE IF EXISTS {self.collection_name}")+ self.conn.commit()++ def col_info(self):+ """+ Get information about a collection.++ Returns:+ Dict[str, Any]: Collection information.+ """+ self.cur.execute(f"""+ SELECT+ table_name,+ (SELECT COUNT(*) FROM {self.collection_name}) as row_count,+ (SELECT pg_size_pretty(pg_total_relation_size('{self.collection_name}'))) as total_size+ FROM information_schema.tables+ WHERE table_schema = 'public' AND table_name = %s+ """, (self.collection_name,))+ result = self.cur.fetchone()+ return {+ "name": result[0],+ "count": result[1],+ "size": result[2]+ }++ def list(self, filters = None, limit = 100):+ """+ List all vectors in a collection.++ Args:+ filters (Dict, optional): Filters to apply to the list.+ limit (int, optional): Number of vectors to return. Defaults to 100.++ Returns:+ List[OutputData]: List of vectors.+ """+ filter_conditions = []+ filter_params = []++ if filters:+ for k, v in filters.items():+ filter_conditions.append(f"payload->>%s = %s")+ filter_params.extend([k, str(v)])++ filter_clause = "WHERE " + " AND ".join(filter_conditions) if filter_conditions else ""++ query = f"""+ SELECT id, vector, payload+ FROM {self.collection_name}+ {filter_clause}+ LIMIT %s+ """++ self.cur.execute(query, (*filter_params, limit))++ results = self.cur.fetchall()+ return [[OutputData(id=str(r[0]), score=None, payload=r[2]) for r in results]]++ def __del__(self):+ """+ Close the database connection when the object is deleted.+ """+ if hasattr(self, 'cur'):+ self.cur.close()+ if hasattr(self, 'conn'):+ self.conn.close()\ No newline at end of filediff --git a/mem0/vector_stores/qdrant.py b/mem0/vector_stores/qdrant.pyindex 396fc776c..f82a602e1 100644--- a/mem0/vector_stores/qdrant.py+++ b/mem0/vector_stores/qdrant.py@@ -61,9 +61,10 @@ class Qdrant(VectorStoreBase):self.client = QdrantClient(**params)- self.create_col(collection_name, embedding_model_dims, on_disk)+ self.collection_name = collection_name+ self.create_col(embedding_model_dims, on_disk)- def create_col(self, name, vector_size, on_disk, distance=Distance.COSINE):+ def create_col(self, vector_size, on_disk, distance=Distance.COSINE):"""Create a new collection.@@ -75,21 +76,20 @@ class Qdrant(VectorStoreBase):# Skip creating collection if already existsresponse = self.list_cols()for collection in response.collections:- if collection.name == name:- logging.debug(f"Collection {name} already exists. Skipping creation.")+ if collection.name == self.collection_name:+ logging.debug(f"Collection {self.collection_name} already exists. Skipping creation.")returnself.client.create_collection(- collection_name=name,+ collection_name=self.collection_name,vectors_config=VectorParams(size=vector_size, distance=distance, on_disk=on_disk),)- def insert(self, name, vectors, payloads=None, ids=None):+ def insert(self, vectors, payloads=None, ids=None):"""Insert vectors into a collection.Args:- name (str): Name of the collection.vectors (list): List of vectors to insert.payloads (list, optional): List of payloads corresponding to vectors. Defaults to None.ids (list, optional): List of IDs corresponding to vectors. Defaults to None.@@ -102,7 +102,7 @@ class Qdrant(VectorStoreBase):)for idx, vector in enumerate(vectors)]- self.client.upsert(collection_name=name, points=points)+ self.client.upsert(collection_name=self.collection_name, points=points)def _create_filter(self, filters):"""@@ -128,12 +128,11 @@ class Qdrant(VectorStoreBase):)return Filter(must=conditions) if conditions else None- def search(self, name, query, limit=5, filters=None):+ def search(self, query, limit=5, filters=None):"""Search for similar vectors.Args:- name (str): Name of the collection.query (list): Query vector.limit (int, optional): Number of results to return. Defaults to 5.filters (dict, optional): Filters to apply to the search. Defaults to None.@@ -143,54 +142,51 @@ class Qdrant(VectorStoreBase):"""query_filter = self._create_filter(filters) if filters else Nonehits = self.client.search(- collection_name=name,+ collection_name=self.collection_name,query_vector=query,query_filter=query_filter,limit=limit,)return hits- def delete(self, name, vector_id):+ def delete(self, vector_id):"""Delete a vector by ID.Args:- name (str): Name of the collection.vector_id (int): ID of the vector to delete."""self.client.delete(- collection_name=name,+ collection_name=self.collection_name,points_selector=PointIdsList(points=[vector_id],),)- def update(self, name, vector_id, vector=None, payload=None):+ def update(self, vector_id, vector=None, payload=None):"""Update a vector and its payload.Args:- name (str): Name of the collection.vector_id (int): ID of the vector to update.vector (list, optional): Updated vector. Defaults to None.payload (dict, optional): Updated payload. Defaults to None."""point = PointStruct(id=vector_id, vector=vector, payload=payload)- self.client.upsert(collection_name=name, points=[point])+ self.client.upsert(collection_name=self.collection_name, points=[point])- def get(self, name, vector_id):+ def get(self, vector_id):"""Retrieve a vector by ID.Args:- name (str): Name of the collection.vector_id (int): ID of the vector to retrieve.Returns:dict: Retrieved vector."""result = self.client.retrieve(- collection_name=name, ids=[vector_id], with_payload=True+ collection_name=self.collection_name, ids=[vector_id], with_payload=True)return result[0] if result else None@@ -203,33 +199,24 @@ class Qdrant(VectorStoreBase):"""return self.client.get_collections()- def delete_col(self, name):- """- Delete a collection.+ def delete_col(self):+ """ Delete a collection. """+ self.client.delete_collection(collection_name=self.collection_name)- Args:- name (str): Name of the collection to delete.- """- self.client.delete_collection(collection_name=name)-- def col_info(self, name):+ def col_info(self):"""Get information about a collection.- Args:- name (str): Name of the collection.-Returns:dict: Collection information."""- return self.client.get_collection(collection_name=name)+ return self.client.get_collection(collection_name=self.collection_name)- def list(self, name, filters=None, limit=100):+ def list(self, filters=None, limit=100):"""List all vectors in a collection.Args:- name (str): Name of the collection.limit (int, optional): Number of vectors to return. Defaults to 100.Returns:@@ -237,7 +224,7 @@ class Qdrant(VectorStoreBase):"""query_filter = self._create_filter(filters) if filters else Noneresult = self.client.scroll(- collection_name=name,+ collection_name=self.collection_name,scroll_filter=query_filter,limit=limit,with_payload=True,