diff --git a/llama_stack/providers/utils/memory/openai_vector_store_mixin.py b/llama_stack/providers/utils/memory/openai_vector_store_mixin.py
index 70bcbba32..4c80d7435 100644
--- a/llama_stack/providers/utils/memory/openai_vector_store_mixin.py
+++ b/llama_stack/providers/utils/memory/openai_vector_store_mixin.py
@@ -353,14 +353,12 @@ class OpenAIVectorStoreMixin(ABC):
         provider_vector_db_id = extra.get("provider_vector_db_id")
         embedding_model = extra.get("embedding_model")
         embedding_dimension = extra.get("embedding_dimension", 384)
-        provider_id = extra.get("provider_id")
+        # use provider_id from router or default to this provider's own ID (need for --stack-config)
+        provider_id = extra.get("provider_id") or getattr(self, "__provider_id__", None)
 
         # Derive the canonical vector_db_id (allow override, else generate)
         vector_db_id = provider_vector_db_id or generate_object_id("vector_store", lambda: f"vs_{uuid.uuid4()}")
 
-        if provider_id is None:
-            raise ValueError("Provider ID is required")
-
         if embedding_model is None:
             raise ValueError("Embedding model is required")
 
@@ -369,6 +367,9 @@ class OpenAIVectorStoreMixin(ABC):
             raise ValueError("Embedding dimension is required")
 
         # Register the VectorDB backing this vector store
+        if provider_id is None:
+            raise ValueError("Provider ID is required but was not provided")
+
         vector_db = VectorDB(
             identifier=vector_db_id,
             embedding_dimension=embedding_dimension,
diff --git a/tests/integration/vector_io/test_vector_io.py b/tests/integration/vector_io/test_vector_io.py
index 53d46e28c..08b082b15 100644
--- a/tests/integration/vector_io/test_vector_io.py
+++ b/tests/integration/vector_io/test_vector_io.py
@@ -224,3 +224,21 @@ def test_query_returns_valid_object_when_identical_to_embedding_in_vdb(
     assert len(response.chunks) > 0
     assert response.chunks[0].metadata["document_id"] == "doc1"
     assert response.chunks[0].metadata["source"] == "precomputed"
+
+
+def test_auto_extract_embedding_dimension(client_with_empty_registry, embedding_model_id):
+    vs = client_with_empty_registry.vector_stores.create(
+        name="test_auto_extract", extra_body={"embedding_model": embedding_model_id}
+    )
+    assert vs.id is not None
+
+
+def test_provider_auto_selection_single_provider(client_with_empty_registry, embedding_model_id):
+    providers = [p for p in client_with_empty_registry.providers.list() if p.api == "vector_io"]
+    if len(providers) != 1:
+        pytest.skip(f"Test requires exactly one vector_io provider, found {len(providers)}")
+
+    vs = client_with_empty_registry.vector_stores.create(
+        name="test_auto_provider", extra_body={"embedding_model": embedding_model_id}
+    )
+    assert vs.id is not None
diff --git a/tests/unit/core/routers/test_vector_io.py b/tests/unit/core/routers/test_vector_io.py
new file mode 100644
index 000000000..997df0d78
--- /dev/null
+++ b/tests/unit/core/routers/test_vector_io.py
@@ -0,0 +1,57 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the terms described in the LICENSE file in
+# the root directory of this source tree.
+
+from unittest.mock import AsyncMock, Mock
+
+import pytest
+
+from llama_stack.apis.vector_io import OpenAICreateVectorStoreRequestWithExtraBody
+from llama_stack.core.routers.vector_io import VectorIORouter
+
+
+async def test_single_provider_auto_selection():
+    # provider_id automatically selected during vector store create() when only one provider available
+    mock_routing_table = Mock()
+    mock_routing_table.impls_by_provider_id = {"inline::faiss": "mock_provider"}
+    mock_routing_table.get_all_with_type = AsyncMock(
+        return_value=[
+            Mock(identifier="all-MiniLM-L6-v2", model_type="embedding", metadata={"embedding_dimension": 384})
+        ]
+    )
+    mock_routing_table.register_vector_db = AsyncMock(
+        return_value=Mock(identifier="vs_123", provider_id="inline::faiss", provider_resource_id="vs_123")
+    )
+    mock_routing_table.get_provider_impl = AsyncMock(
+        return_value=Mock(openai_create_vector_store=AsyncMock(return_value=Mock(id="vs_123")))
+    )
+    router = VectorIORouter(mock_routing_table)
+    request = OpenAICreateVectorStoreRequestWithExtraBody.model_validate(
+        {"name": "test_store", "embedding_model": "all-MiniLM-L6-v2"}
+    )
+
+    result = await router.openai_create_vector_store(request)
+    assert result.id == "vs_123"
+
+
+async def test_create_vector_stores_multiple_providers_missing_provider_id_error():
+    # if multiple providers are available, vector store create will error without provider_id
+    mock_routing_table = Mock()
+    mock_routing_table.impls_by_provider_id = {
+        "inline::faiss": "mock_provider_1",
+        "inline::sqlite-vec": "mock_provider_2",
+    }
+    mock_routing_table.get_all_with_type = AsyncMock(
+        return_value=[
+            Mock(identifier="all-MiniLM-L6-v2", model_type="embedding", metadata={"embedding_dimension": 384})
+        ]
+    )
+    router = VectorIORouter(mock_routing_table)
+    request = OpenAICreateVectorStoreRequestWithExtraBody.model_validate(
+        {"name": "test_store", "embedding_model": "all-MiniLM-L6-v2"}
+    )
+
+    with pytest.raises(ValueError, match="Multiple vector_io providers available"):
+        await router.openai_create_vector_store(request)