microsoft/semantic-kernel · error · VectorStoreInitializationException

Failed to create Azure AI Search settings.

Error message

Failed to create Azure AI Search settings.

What it means

Raised in the AzureAISearchStore constructor when no search_index_client is supplied and AzureAISearchSettings construction raises a pydantic ValidationError. The store-level settings require a valid HTTPS endpoint (HttpsUrl). The original ValidationError is chained via 'from exc'. This is the store-level analogue of errors 1224/1225 (which occur at the collection level).

Source

Thrown at python/semantic_kernel/connectors/azure_ai_search.py:783

        search_index_client: SearchIndexClient | None = None,
        embedding_generator: "EmbeddingGeneratorBase | None" = None,
        env_file_path: str | None = None,
        env_file_encoding: str | None = None,
    ) -> None:
        """Initializes a new instance of the AzureAISearchStore class."""
        managed_client: bool = False
        endpoint: str | None = None
        credential: AzureKeyCredential | AsyncTokenCredential | None = None
        if not search_index_client:
            try:
                azure_ai_search_settings = AzureAISearchSettings(
                    env_file_path=env_file_path,
                    endpoint=search_endpoint,
                    api_key=api_key,
                    env_file_encoding=env_file_encoding,
                )
            except ValidationError as exc:
                raise VectorStoreInitializationException("Failed to create Azure AI Search settings.") from exc
            endpoint = str(azure_ai_search_settings.endpoint)
            credential = _resolve_credential(
                azure_ai_search_settings,
                azure_credential=azure_credentials,
                token_credential=token_credentials,
            )
            search_index_client = _get_search_index_client(
                azure_ai_search_settings=azure_ai_search_settings,
                azure_credential=azure_credentials,
                token_credential=token_credentials,
            )
            managed_client = True
        else:
            endpoint = search_endpoint
            credential = azure_credentials or token_credentials or (AzureKeyCredential(api_key) if api_key else None)

        super().__init__(
            search_index_client=search_index_client,

View on GitHub (pinned to c028a0c7dc)

Solutions

1. Pass search_endpoint='https://<svc>.search.windows.net' explicitly, or set AZURE_AI_SEARCH_ENDPOINT.
2. Read exc.__cause__ (the ValidationError) to identify the exact failing field and reason.
3. Ensure the endpoint uses the https:// scheme (HttpUrl/HttpsUrl rejects http://).
4. Verify env_file_path and env_file_encoding if loading from a .env file.

Example fix

// before
store = AzureAISearchStore()  # AZURE_AI_SEARCH_ENDPOINT not set

// after
store = AzureAISearchStore(
    search_endpoint="https://mysvc.search.windows.net",
    api_key="<key>",
)

import os

def require_store_endpoint(search_endpoint=None) -> str:
    ep = search_endpoint or os.getenv("AZURE_AI_SEARCH_ENDPOINT")
    if not ep or not ep.startswith("https://"):
        raise ValueError("AZURE_AI_SEARCH_ENDPOINT (https://) required for AzureAISearchStore")
    return ep

require_store_endpoint()

from semantic_kernel.exceptions import VectorStoreInitializationException
try:
    store = AzureAISearchStore()
except VectorStoreInitializationException as e:
    cause = e.__cause__  # pydantic ValidationError
    for err in cause.errors():
        print(err["loc"], err["msg"])
    raise

When it happens

Trigger: Constructing AzureAISearchStore() with AZURE_AI_SEARCH_ENDPOINT unset and no search_endpoint argument; endpoint not a valid HTTPS URL; or env_file_encoding misconfigured so the .env fails to load. Since no client is provided, settings must be built from kwargs/env, and the endpoint requirement triggers the failure.

Common situations: First-time initialization without environment configuration; endpoint typo; running in a fresh environment/CI where AZURE_AI_SEARCH_ENDPOINT was not injected; wrong .env path.

Related errors

• Failed to create Azure Cognitive Search settings.
• Error: missing Azure AI Search client credentials.
• OpenAI/Azure OpenAI configuration was not found.
• Could not find configuration section {caller}
• AI Chat Service type '{appConfig.RagConfig.AIChatService}' i

AI-assisted analysis of microsoft/semantic-kernel@c028a0c7dc (2026-08-13). Data as JSON: /api/errors/e0d83ff61eb0315e. Report an issue: GitHub.