feat: Add openAI compatible APIs to Qdrant (#2465)

# What does this PR do?
Adds support to Vector store Open AI APIs in Qdrant.

<!-- If resolving an issue, uncomment and update the line below -->
 Closes #2463 


## Test Plan
<!-- Describe the tests you ran to verify your changes with result
summaries. *Provide clear instructions so the plan can be easily
re-executed.* -->

Signed-off-by: Varsha Prasad Narsing <varshaprasad96@gmail.com>
Co-authored-by: ehhuang <ehhuang@users.noreply.github.com>
Co-authored-by: Francisco Arceo <arceofrancisco@gmail.com>

tests/integration/vector_io/test_openai_vector_stores.py

@@ -29,6 +29,8 @@ def skip_if_provider_doesnt_support_openai_vector_stores(client_with_models):
             "inline::chromadb",
             "remote::pgvector",
             "remote::chromadb",
+            "remote::qdrant",
+            "inline::qdrant",
             "remote::weaviate",
         ]:
             return

tests/integration/vector_io/test_vector_io.py

@@ -125,6 +125,8 @@ def test_insert_chunks(client_with_empty_registry, embedding_model_id, embedding
 def test_insert_chunks_with_precomputed_embeddings(client_with_empty_registry, embedding_model_id, embedding_dimension):
     vector_io_provider_params_dict = {
         "inline::milvus": {"score_threshold": -1.0},
+        "remote::qdrant": {"score_threshold": -1.0},
+        "inline::qdrant": {"score_threshold": -1.0},
     }
     vector_db_id = "test_precomputed_embeddings_db"
     client_with_empty_registry.vector_dbs.register(
@@ -168,6 +170,8 @@ def test_query_returns_valid_object_when_identical_to_embedding_in_vdb(
 ):
     vector_io_provider_params_dict = {
         "inline::milvus": {"score_threshold": 0.0},
+        "remote::qdrant": {"score_threshold": 0.0},
+        "inline::qdrant": {"score_threshold": 0.0},
     }
     vector_db_id = "test_precomputed_embeddings_db"
     client_with_empty_registry.vector_dbs.register(

tests/unit/providers/vector_io/conftest.py

@@ -17,10 +17,12 @@ from llama_stack.providers.inline.vector_io.chroma.config import ChromaVectorIOC
 from llama_stack.providers.inline.vector_io.faiss.config import FaissVectorIOConfig
 from llama_stack.providers.inline.vector_io.faiss.faiss import FaissIndex, FaissVectorIOAdapter
 from llama_stack.providers.inline.vector_io.milvus.config import MilvusVectorIOConfig, SqliteKVStoreConfig
+from llama_stack.providers.inline.vector_io.qdrant import QdrantVectorIOConfig
 from llama_stack.providers.inline.vector_io.sqlite_vec import SQLiteVectorIOConfig
 from llama_stack.providers.inline.vector_io.sqlite_vec.sqlite_vec import SQLiteVecIndex, SQLiteVecVectorIOAdapter
 from llama_stack.providers.remote.vector_io.chroma.chroma import ChromaIndex, ChromaVectorIOAdapter, maybe_await
 from llama_stack.providers.remote.vector_io.milvus.milvus import MilvusIndex, MilvusVectorIOAdapter
+from llama_stack.providers.remote.vector_io.qdrant.qdrant import QdrantVectorIOAdapter
 
 EMBEDDING_DIMENSION = 384
 COLLECTION_PREFIX = "test_collection"
@@ -280,6 +282,57 @@ async def chroma_vec_adapter(chroma_vec_db_path, mock_inference_api, embedding_d
     await adapter.shutdown()
 
 
+@pytest.fixture
+def qdrant_vec_db_path(tmp_path_factory):
+    import uuid
+
+    db_path = str(tmp_path_factory.getbasetemp() / f"test_qdrant_{uuid.uuid4()}.db")
+    return db_path
+
+
+@pytest.fixture
+async def qdrant_vec_adapter(qdrant_vec_db_path, mock_inference_api, embedding_dimension):
+    import uuid
+
+    config = QdrantVectorIOConfig(
+        db_path=qdrant_vec_db_path,
+        kvstore=SqliteKVStoreConfig(),
+    )
+    adapter = QdrantVectorIOAdapter(
+        config=config,
+        inference_api=mock_inference_api,
+        files_api=None,
+    )
+    collection_id = f"qdrant_test_collection_{uuid.uuid4()}"
+    await adapter.initialize()
+    await adapter.register_vector_db(
+        VectorDB(
+            identifier=collection_id,
+            provider_id="test_provider",
+            embedding_model="test_model",
+            embedding_dimension=embedding_dimension,
+        )
+    )
+    adapter.test_collection_id = collection_id
+    yield adapter
+    await adapter.shutdown()
+
+
+@pytest.fixture
+async def qdrant_vec_index(qdrant_vec_db_path, embedding_dimension):
+    import uuid
+
+    from qdrant_client import AsyncQdrantClient
+
+    from llama_stack.providers.remote.vector_io.qdrant.qdrant import QdrantIndex
+
+    client = AsyncQdrantClient(path=qdrant_vec_db_path)
+    collection_name = f"qdrant_test_collection_{uuid.uuid4()}"
+    index = QdrantIndex(client, collection_name)
+    yield index
+    await index.delete()
+
+
 @pytest.fixture
 def vector_io_adapter(vector_provider, request):
     """Returns the appropriate vector IO adapter based on the provider parameter."""
@@ -288,6 +341,7 @@ def vector_io_adapter(vector_provider, request):
         "faiss": "faiss_vec_adapter",
         "sqlite_vec": "sqlite_vec_adapter",
         "chroma": "chroma_vec_adapter",
+        "qdrant": "qdrant_vec_adapter",
     }
     return request.getfixturevalue(vector_provider_dict[vector_provider])
 

tests/unit/providers/vector_io/test_qdrant.py

@@ -23,6 +23,7 @@ from llama_stack.providers.inline.vector_io.qdrant.config import (
 from llama_stack.providers.remote.vector_io.qdrant.qdrant import (
     QdrantVectorIOAdapter,
 )
+from llama_stack.providers.utils.kvstore.config import SqliteKVStoreConfig
 
 # This test is a unit test for the QdrantVectorIOAdapter class. This should only contain
 # tests which are specific to this class. More general (API-level) tests should be placed in
@@ -36,7 +37,8 @@ from llama_stack.providers.remote.vector_io.qdrant.qdrant import (
 
 @pytest.fixture
 def qdrant_config(tmp_path) -> InlineQdrantVectorIOConfig:
-    return InlineQdrantVectorIOConfig(path=os.path.join(tmp_path, "qdrant.db"))
+    kvstore_config = SqliteKVStoreConfig(db_name=os.path.join(tmp_path, "test_kvstore.db"))
+    return InlineQdrantVectorIOConfig(path=os.path.join(tmp_path, "qdrant.db"), kvstore=kvstore_config)
 
 
 @pytest.fixture(scope="session")
@@ -50,6 +52,9 @@ def mock_vector_db(vector_db_id) -> MagicMock:
     mock_vector_db.embedding_model = "embedding_model"
     mock_vector_db.identifier = vector_db_id
     mock_vector_db.embedding_dimension = 384
+    mock_vector_db.model_dump_json.return_value = (
+        '{"identifier": "' + vector_db_id + '", "embedding_model": "embedding_model", "embedding_dimension": 384}'
+    )
     return mock_vector_db
 
 
@@ -69,7 +74,7 @@ def mock_api_service(sample_embeddings):
 
 @pytest.fixture
 async def qdrant_adapter(qdrant_config, mock_vector_db_store, mock_api_service, loop) -> QdrantVectorIOAdapter:
-    adapter = QdrantVectorIOAdapter(config=qdrant_config, inference_api=mock_api_service)
+    adapter = QdrantVectorIOAdapter(config=qdrant_config, inference_api=mock_api_service, files_api=None)
     adapter.vector_db_store = mock_vector_db_store
     await adapter.initialize()
     yield adapter