🐛 Describe the bug
Summary
I am just testing around with self-hosted Mem0, with various vector DBs.
I would love to try filtering but having a hard time making it work.
The Qdrant did not work for the same reason descried in another issue.
When using Weaviate, it throws no error but any filter seems to be ignored or not working.
Official mem0 doc says:
Full operator coverage with advanced text filters. Best option when you need hybrid text + metadata queries.
Environment
- mem0 version: 1.0.3
- weaviate-client version: 4.19.2
- weaviate server version: semitechnologies/weaviate: latest (Docker image), 1.36.0-rc.0
- Python version: 3.12.11
- OS: MacOs 15.5
Weaviate docker compose
weaviate:
image: semitechnologies/weaviate:latest
ports:
- "8080:8080"
- "50051:50051"
environment:
QUERY_DEFAULTS_LIMIT: 25
AUTHENTICATION_ANONYMOUS_ACCESS_ENABLED: "true"
PERSISTENCE_DATA_PATH: "/var/lib/weaviate"
DEFAULT_VECTORIZER_MODULE: "none"
ENABLE_MODULES: ""Sample python code
Config
from mem0 import Memory
config = {
"vector_store": {
"provider": "weaviate",
"config": {
"collection_name": "test",
"cluster_url": "http://localhost:8080",
"auth_client_secret": None,
}
}
}
m = Memory.from_config(config)By the way it spits out
DeprecationWarning: Dep024: You are using the vectorizer_config argument in collection.config.create(), which is deprecated. Use the vector_config argument instead.
Memory Add
messages = [
{"role": "user", "content": "I'm planning to watch a movie tonight. Any recommendations?"},
{"role": "assistant", "content": "How about thriller movies? They can be quite engaging."},
{"role": "user", "content": "I’m not a big fan of thriller movies but I love sci-fi movies."},
{"role": "assistant", "content": "Got it! I'll avoid thriller recommendations and suggest sci-fi movies in the future."}
]
m.add(messages, user_id="alice", metadata={"category": "movies", "test": 0})Seems properly saving something.
{'results': [{'id': '1fb2ee17-57f0-4f95-8c81-df2a645b9bec',
'memory': 'Loves sci-fi movies',
'event': 'ADD'},
{'id': '2b920e65-18a0-4abd-b9a6-4f9c2de8302a',
'memory': 'Not a big fan of thriller movies',
'event': 'ADD'}]}
Searching
Not-Working patterns
import json
results = m.search(
"recommend content",
user_id="alice",
filters={"category": "test"}
)
print(json.dumps(results, indent=2))
print(f"Search records: {len(results["results"])}")
Since filters specifies category as "test", I expected no results to be returned but got:
{
"results": [
{
"id": "1fb2ee17-57f0-4f95-8c81-df2a645b9bec",
"memory": "Loves sci-fi movies",
"hash": "1110b1af77367917ea2022355a16f187",
"metadata": {
"category": "movies"
},
"score": 0.699999988079071,
"created_at": "2026-02-15T01:59:48.810746-08:00",
"updated_at": null,
"user_id": "alice"
},
{
"id": "2b920e65-18a0-4abd-b9a6-4f9c2de8302a",
"memory": "Not a big fan of thriller movies",
"hash": "028dfab4483f28980e292f62578d3293",
"metadata": {
"category": "movies"
},
"score": 0.0,
"created_at": "2026-02-15T01:59:49.856939-08:00",
"updated_at": null,
"user_id": "alice"
}
]
}
Search records: 2
Any other filters seems not working at all.
Working patterns
Pattern 1
import json
results = m.search(
"recommend content",
user_id="alice",
)
print(json.dumps(results, indent=2))
print(f"Search records: {len(results["results"])}")
Output is:
import json
results = m.search(
"recommend content",
user_id="alice",
)
print(json.dumps(results, indent=2))
print(f"Search records: {len(results["results"])}")
Pattern 2
import json
results = m.search(
"recommend content",
user_id="bob",
)
print(json.dumps(results, indent=2))
print(f"Search records: {len(results["results"])}")
Output is:
{
"results": []
}
Search records: 0
Pattern 3
import json
results = m.search(
"recommend content",
user_id="alice",
filters={"category": "movies"}
)
print(json.dumps(results, indent=2))
print(f"Search records: {len(results["results"])}")
Output is:
(This looks like it is working, but I suppose the filter is simply ignored.)
{
"results": [
{
"id": "1fb2ee17-57f0-4f95-8c81-df2a645b9bec",
"memory": "Loves sci-fi movies",
"hash": "1110b1af77367917ea2022355a16f187",
"metadata": {
"category": "movies"
},
"score": 0.699999988079071,
"created_at": "2026-02-15T01:59:48.810746-08:00",
"updated_at": null,
"user_id": "alice"
},
{
"id": "2b920e65-18a0-4abd-b9a6-4f9c2de8302a",
"memory": "Not a big fan of thriller movies",
"hash": "028dfab4483f28980e292f62578d3293",
"metadata": {
"category": "movies"
},
"score": 0.0,
"created_at": "2026-02-15T01:59:49.856939-08:00",
"updated_at": null,
"user_id": "alice"
}
]
}
Search records: 2
🐛 Describe the bug
Summary
I am just testing around with self-hosted Mem0, with various vector DBs.
I would love to try filtering but having a hard time making it work.
The Qdrant did not work for the same reason descried in another issue.
When using Weaviate, it throws no error but any filter seems to be ignored or not working.
Official mem0 doc says:
Environment
Weaviate docker compose
Sample python code
Config
By the way it spits out
Memory Add
Seems properly saving something.
Searching
Not-Working patterns
Since filters specifies category as "test", I expected no results to be returned but got:
Any other filters seems not working at all.
Working patterns
Pattern 1
Output is:
Pattern 2
Output is:
Pattern 3
Output is:
(This looks like it is working, but I suppose the filter is simply ignored.)