What causes "{field.index_kind} not supported in Azure AI Search." in microsoft/semantic-kernel?

Set the vector field's index_kind to IndexKind.HNSW (recommended for most workloads), IndexKind.FLAT, or IndexKind.DEFAULT.

microsoft/semantic-kernel · error · VectorStoreOperationException

{field.index_kind} not supported in Azure AI Search.

Error message

{field.index_kind} not supported in Azure AI Search.

What it means

Raised in _definition_to_azure_ai_search_index when a VECTOR field's index_kind is not a key in INDEX_ALGORITHM_MAP. Azure AI Search only supports HNSW, FLAT (exhaustive KNN), and DEFAULT (alias for HNSW). Other index kinds defined by the IndexKind enum (IVF_FLAT, DISK_ANN, QUANTIZED_FLAT, DYNAMIC) are valid for other stores but not for Azure AI Search, so index creation fails with a VectorStoreOperationException.

Source

Thrown at python/semantic_kernel/connectors/azure_ai_search.py:261

                    sortable=not type_.startswith("Collection") or type_ == "Edm.ComplexType",
                    hidden=False,
                )
            )
        elif field.field_type == FieldTypes.KEY:
            fields.append(
                SimpleField(
                    name=field.storage_name or field.name,
                    type="Edm.String",  # hardcoded, only allowed type for key
                    key=True,
                    filterable=True,
                    searchable=True,
                )
            )
        elif field.field_type == FieldTypes.VECTOR:
            if not field.type_:
                logger.debug(f"Field {field.name} has not specified type, defaulting to Collection(Edm.Single).")
            if field.index_kind not in INDEX_ALGORITHM_MAP:
                raise VectorStoreOperationException(f"{field.index_kind} not supported in Azure AI Search.")
            if field.distance_function not in DISTANCE_FUNCTION_MAP:
                raise VectorStoreOperationException(f"{field.distance_function} not supported in Azure AI Search.")

            profile_name = f"{field.storage_name or field.name}_profile"
            algo_name = f"{field.storage_name or field.name}_algorithm"
            fields.append(
                SearchField(
                    name=field.storage_name or field.name,
                    type=TYPE_MAP_VECTOR[field.type_ or "default"],
                    searchable=True,
                    vector_search_dimensions=field.dimensions,
                    vector_search_profile_name=profile_name,
                    hidden=False,
                )
            )
            search_profiles.append(
                VectorSearchProfile(
                    name=profile_name,

View on GitHub (pinned to c028a0c7dc)

Solutions

Set the vector field's index_kind to IndexKind.HNSW (recommended for most workloads), IndexKind.FLAT, or IndexKind.DEFAULT.
If you need disk-based or quantized indexes, use Azure Cosmos DB NoSQL or MongoDB Atlas instead of Azure AI Search.
Validate all vector fields in the definition against INDEX_ALGORITHM_MAP before calling ensure_collection_exists().

Example fix

// before
field(type_='float', name='embedding', index_kind=IndexKind.DISK_ANN)

// after
field(type_='float', name='embedding', index_kind=IndexKind.HNSW)

Defensive patterns

Strategy: validation

Validate before calling

from semantic_kernel.connectors.azure_ai_search import INDEX_ALGORITHM_MAP

def validate_vector_index_kinds(definition) -> list[str]:
    bad = []
    for f in definition.fields:
        if f.field_type.value == "vector" and f.index_kind not in INDEX_ALGORITHM_MAP:
            bad.append(f"{f.name}: {f.index_kind}")
    return bad

assert not validate_vector_index_kinds(definition)

Try / catch

from semantic_kernel.exceptions import VectorStoreOperationException
try:
    await collection.ensure_collection_exists()
except VectorStoreOperationException as e:
    if "not supported in Azure AI Search" in str(e) and "index" in str(e).lower():
        # change index_kind to HNSW/FLAT/DEFAULT
        ...
    raise

Prevention

Default vector fields to IndexKind.HNSW unless you need exact (FLAT) search.
When porting a definition between stores, re-map index_kind per store — each has its own MAP.
Validate the definition against INDEX_ALGORITHM_MAP in a test.

When it happens

Trigger: Calling ensure_collection_exists() on a collection whose definition sets a vector field's index_kind to IndexKind.IVF_FLAT, IndexKind.DISK_ANN, IndexKind.QUANTIZED_FLAT, or IndexKind.DYNAMIC. Occurs at index-build time during collection creation.

Common situations: Porting a model definition from Azure Cosmos DB NoSQL or MongoDB Atlas (which accept disk_ann / quantized_flat / ivf_flat) to Azure AI Search without changing index_kind; copy-pasting a definition from a cross-store tutorial.

Related errors

AI-assisted analysis of microsoft/semantic-kernel@c028a0c7dc (2026-08-13). Data as JSON: /api/errors/1256d975388ef10a. Report an issue: GitHub.