azure-search-documents_12.1.0

12.1.0 (2026-11-13)

Features Added

• Added support for the 2026-10-01 GA API, which is now the default.
• Below enum members and parameters are added for API version selection and document search:
  • azure.search.documents.ApiVersion.V2026_10_01
  • azure.search.documents.SearchClient.get_document.enable_elevated_read
  • azure.search.documents.SearchClient.get_document.query_source_authorization
  • azure.search.documents.SearchClient.search.enable_elevated_read
  • azure.search.documents.SearchClient.search.more_like_this
  • azure.search.documents.SearchClient.search.query_source_authorization
• Below pageable operations and enum members are added for resource listing:
  • azure.search.documents.indexes.SearchIndexClient.list_index_stats_summary
  • azure.search.documents.indexes.SearchIndexClient.list_synonym_maps
  • azure.search.documents.indexes.SearchIndexerClient.list_data_source_connections
  • azure.search.documents.indexes.SearchIndexerClient.list_indexers
  • azure.search.documents.indexes.SearchIndexerClient.list_skillsets
  • azure.search.documents.indexes.models.ListingSearchType.PREFIX
• Added search, page_size, and search_type controls to resource listings, including the existing
  index, alias, synonym map, knowledge base, knowledge source, indexer, data source, and skillset
  listing helpers. The existing get_* helpers continue to return lists.
• Below models and properties are added for index statistics and indexer runtime reporting:
  • azure.search.documents.indexes.models.IndexStatisticsSummary
  • azure.search.documents.indexes.models.IndexerRuntime
  • azure.search.documents.indexes.models.SearchIndexerStatus.runtime
  • azure.search.documents.indexes.models.SearchServiceStatistics.indexers_runtime
  • azure.search.documents.indexes.models.ServiceIndexersRuntime
• Added azure.search.documents.indexes.SearchIndexerClient.resync, using IndexerResyncBody and
  azure.search.documents.indexes.models.IndexerResyncOption.PERMISSIONS, to resynchronize indexed
  permission data.
• Below models and operations are added for File knowledge sources:
  • azure.search.documents.indexes.SearchIndexClient.delete_knowledge_source_file
  • azure.search.documents.indexes.SearchIndexClient.list_knowledge_source_files
  • azure.search.documents.indexes.SearchIndexClient.update_knowledge_source_file
  • azure.search.documents.indexes.SearchIndexClient.upload_knowledge_source_file_multipart
  • azure.search.documents.indexes.models.FileKnowledgeSource
  • azure.search.documents.indexes.models.FileKnowledgeSourceExtractionMode
  • azure.search.documents.indexes.models.FileKnowledgeSourceParameters
  • azure.search.documents.indexes.models.FileUploadMetadata
  • azure.search.documents.indexes.models.KnowledgeSourceFile
  • azure.search.documents.indexes.models.KnowledgeSourceFileCapacity
  • azure.search.documents.indexes.models.UpdateKnowledgeSourceFileRequest
  • azure.search.documents.indexes.models.UploadKnowledgeSourceFileMultipartRequest
  • azure.search.documents.knowledgebases.models.FileKnowledgeSourceParams
• Added file metadata, relative-path prefixes, parsing and extraction modes, timestamps, file sizes,
  and processing errors to the File knowledge source lifecycle. File listing supports prefix
  filtering and pagination. azure.search.documents.knowledgebases.models.KnowledgeSourceStatus.file_capacity
  reports maximum file size, maximum file count, and remaining file count.
• Below models are added for indexed SQL and SharePoint knowledge sources:
  • azure.search.documents.indexes.models.ContentColumnMapping
  • azure.search.documents.indexes.models.EmbeddingColumnMapping
  • azure.search.documents.indexes.models.IndexedSharePointContainerName
  • azure.search.documents.indexes.models.IndexedSharePointKnowledgeSource
  • azure.search.documents.indexes.models.IndexedSharePointKnowledgeSourceParameters
  • azure.search.documents.indexes.models.IndexedSqlKnowledgeSource
  • azure.search.documents.indexes.models.IndexedSqlKnowledgeSourceParameters
  • azure.search.documents.knowledgebases.models.IndexedSharePointKnowledgeSourceParams
  • azure.search.documents.knowledgebases.models.IndexedSqlKnowledgeSourceParams
• Below models and properties are added for document permissions and private ingestion:
  • azure.search.documents.indexes.models.IndexingParametersConfiguration.refresh_all_acls
  • azure.search.documents.indexes.models.IndexerPermissionOption
  • azure.search.documents.indexes.models.KnowledgeSourceIngestionPermissionOption
  • azure.search.documents.indexes.models.PermissionFilter
  • azure.search.documents.indexes.models.SearchField.organization_access_expiration
  • azure.search.documents.indexes.models.SearchField.permission_filter
  • azure.search.documents.indexes.models.SearchField.sharepoint_site_url
  • azure.search.documents.indexes.models.SearchIndex.permission_filter_option
  • azure.search.documents.indexes.models.SearchIndex.share_point_connector_app_registration
  • azure.search.documents.indexes.models.SearchIndexPermissionFilterOption
  • azure.search.documents.indexes.models.SearchIndexerDataSourceConnection.indexer_permission_options
  • azure.search.documents.indexes.models.SearchIndexerDataUserAssignedIdentity.federated_identity_client_id
  • azure.search.documents.indexes.models.SharePointConnectorAppRegistration
  • azure.search.documents.knowledgebases.models.KnowledgeSourceIngestionParameters.ingestion_permission_options
  • azure.search.documents.knowledgebases.models.KnowledgeSourceIngestionParameters.network_access_mode
  • azure.search.documents.knowledgebases.models.KnowledgeSourceNetworkAccessMode
• Added azure.search.documents.knowledgebases.KnowledgeBaseRetrievalClient.retrieve.query_source_authorization
  to forward the querying user's authorization to knowledge retrieval.
• Below properties are added for search-index and web knowledge source configuration:
  • azure.search.documents.indexes.models.SearchIndexKnowledgeSourceParameters.base_filter
  • azure.search.documents.indexes.models.WebKnowledgeSourceParameters.count
  • azure.search.documents.indexes.models.WebKnowledgeSourceParameters.freshness
  • azure.search.documents.indexes.models.WebKnowledgeSourceParameters.language
  • azure.search.documents.indexes.models.WebKnowledgeSourceParameters.market
• Below models and properties are added for knowledge base configuration and retrieval requests:
  • azure.search.documents.indexes.models.KnowledgeBase.answer_instructions
  • azure.search.documents.indexes.models.KnowledgeBase.cors_options
  • azure.search.documents.indexes.models.KnowledgeBase.output_mode
  • azure.search.documents.indexes.models.KnowledgeBase.retrieval_instructions
  • azure.search.documents.indexes.models.KnowledgeBase.retrieval_reasoning_effort
  • azure.search.documents.knowledgebases.models.KnowledgeBaseRetrievalRequest.max_output_documents
  • azure.search.documents.knowledgebases.models.KnowledgeBaseRetrievalRequest.messages
  • azure.search.documents.knowledgebases.models.KnowledgeBaseRetrievalRequest.output_mode
  • azure.search.documents.knowledgebases.models.KnowledgeBaseRetrievalRequest.retrieval_reasoning_effort
  • azure.search.documents.knowledgebases.models.KnowledgeRetrievalLowReasoningEffort
  • azure.search.documents.knowledgebases.models.KnowledgeRetrievalMediumReasoningEffort
  • azure.search.documents.knowledgebases.models.KnowledgeRetrievalOutputMode
• Added per-source retrieval controls through
  azure.search.documents.knowledgebases.models.KnowledgeSourceParams.always_query_source,
  azure.search.documents.knowledgebases.models.KnowledgeSourceParams.fail_on_error, and
  azure.search.documents.knowledgebases.models.KnowledgeSourceParams.max_output_documents.
• Below activity, argument, reference, and diagnostic models are added for retrieval responses:
  • azure.search.documents.knowledgebases.models.KnowledgeBaseActivityRecordModel
  • azure.search.documents.knowledgebases.models.KnowledgeBaseAzureBlobActivityArguments
  • azure.search.documents.knowledgebases.models.KnowledgeBaseAzureBlobActivityRecord
  • azure.search.documents.knowledgebases.models.KnowledgeBaseFileActivityArguments
  • azure.search.documents.knowledgebases.models.KnowledgeBaseFileActivityRecord
  • azure.search.documents.knowledgebases.models.KnowledgeBaseFileReference
  • azure.search.documents.knowledgebases.models.KnowledgeBaseIndexedOneLakeActivityArguments
  • azure.search.documents.knowledgebases.models.KnowledgeBaseIndexedOneLakeActivityRecord
  • azure.search.documents.knowledgebases.models.KnowledgeBaseIndexedSharePointActivityArguments
  • azure.search.documents.knowledgebases.models.KnowledgeBaseIndexedSharePointActivityRecord
  • azure.search.documents.knowledgebases.models.KnowledgeBaseIndexedSharePointReference
  • azure.search.documents.knowledgebases.models.KnowledgeBaseIndexedSqlActivityArguments
  • azure.search.documents.knowledgebases.models.KnowledgeBaseIndexedSqlActivityRecord
  • azure.search.documents.knowledgebases.models.KnowledgeBaseIndexedSqlReference
  • azure.search.documents.knowledgebases.models.KnowledgeBaseModelAnswerSynthesisActivityRecord
  • azure.search.documents.knowledgebases.models.KnowledgeBaseModelQueryPlanningActivityRecord
  • azure.search.documents.knowledgebases.models.KnowledgeBaseSearchIndexActivityArguments
  • azure.search.documents.knowledgebases.models.KnowledgeBaseSearchIndexActivityRecord
  • azure.search.documents.knowledgebases.models.KnowledgeBaseWebActivityArguments
  • azure.search.documents.knowledgebases.models.KnowledgeBaseWebActivityRecord
• Added started_at, completed_at, and warning to retrieval activity records, model and
  deployment details to model activities, and citation_url to supported knowledge base references.
• Added azure.search.documents.knowledgebases.KnowledgeBaseRetrievalClient.retrieve_stream for
  native typed server-sent event streaming. Iteration yields retrieval-start, activity-start,
  completed-activity, completed-answer, reference-list, stream-error, and response-completion
  payloads. Completion payloads distinguish successful and partial results with status 200 or 206.
• Below payload models and status enums are added for retrieval streaming:
  • azure.search.documents.knowledgebases.models.KnowledgeBaseActivityStartedEvent
  • azure.search.documents.knowledgebases.models.KnowledgeBaseAnswerCompletedEvent
  • azure.search.documents.knowledgebases.models.KnowledgeBaseResponseCompletedEvent
  • azure.search.documents.knowledgebases.models.KnowledgeBaseRetrievalStartedEvent
  • azure.search.documents.knowledgebases.models.KnowledgeBaseRetrievalStatusCode
  • azure.search.documents.knowledgebases.models.KnowledgeBaseStreamErrorEvent
• Below generated stream envelope models are added. Stream iteration returns their payloads
  directly, rather than envelope objects:
  • azure.search.documents.knowledgebases.models.KnowledgeBaseActivityCompletedStreamEvent
  • azure.search.documents.knowledgebases.models.KnowledgeBaseActivityStartedStreamEvent
  • azure.search.documents.knowledgebases.models.KnowledgeBaseAnswerCompletedStreamEvent
  • azure.search.documents.knowledgebases.models.KnowledgeBaseErrorStreamEvent
  • azure.search.documents.knowledgebases.models.KnowledgeBaseReferencesCompletedStreamEvent
  • azure.search.documents.knowledgebases.models.KnowledgeBaseResponseCompletedStreamEvent
  • azure.search.documents.knowledgebases.models.KnowledgeBaseRetrievalStartedStreamEvent
  • azure.search.documents.knowledgebases.models.KnowledgeBaseRetrievalStreamEvent
• Below models, properties, and enum members are added for Content Understanding
  (#48555):
  • azure.search.documents.indexes.models.ContentUnderstandingSkill.model_deployment
  • azure.search.documents.indexes.models.ContentUnderstandingSkill.model_name
  • azure.search.documents.indexes.models.ContentUnderstandingSkillChunkingMethod
  • azure.search.documents.indexes.models.ContentUnderstandingSkillChunkingProperties.method
  • azure.search.documents.indexes.models.ContentUnderstandingSkillChunkingUnit.TOKENS
• Below enum members are added for supported Azure OpenAI model names:
  • azure.search.documents.indexes.models.AzureOpenAIModelName.GPT54
  • azure.search.documents.indexes.models.AzureOpenAIModelName.GPT55
  • azure.search.documents.indexes.models.AzureOpenAIModelName.GPT56_LUNA
  • azure.search.documents.indexes.models.AzureOpenAIModelName.GPT56_SOL
  • azure.search.documents.indexes.models.AzureOpenAIModelName.GPT56_TERRA
• Below enum names are added for already supported Azure OpenAI model values. The existing
  underscored names remain available as aliases:
  • azure.search.documents.indexes.models.AzureOpenAIModelName.GPT5_4_MINI
  • azure.search.documents.indexes.models.AzureOpenAIModelName.GPT5_4_NANO
  • azure.search.documents.indexes.models.AzureOpenAIModelName.GPT5_MINI
  • azure.search.documents.indexes.models.AzureOpenAIModelName.GPT5_NANO
• Added public input TypedDicts and union aliases under azure.search.documents.types,
  azure.search.documents.indexes.types, and azure.search.documents.knowledgebases.types.
  Supported operation and customized-constructor inputs accept these types alongside existing
  models and mappings; vector queries accept sequences of model or input-query objects.
• Added asynchronous equivalents of the new operations.

Bugs Fixed

• Fixed search metadata accessors clearing continuation tokens or skipping a prefetched first page.
  Counts, facets, coverage, and semantic answers remain available after result iteration.
• Fixed search continuation tokens using a hardcoded API version instead of the client's configured
  version.
• Fixed semantic answer and caption options serializing enum names instead of wire values, and
  threshold-only answer options using an incorrect separator
  (#40497).
• Improved large document upload and merge performance by avoiding redundant serialization
  (#46860).
• Made SearchFieldDataType.Collection visible to static type checkers
  (#47929).
• Fixed IndexDocumentsBatch action replacement, dequeue, and enqueue sending stale document
  actions instead of the batch's current contents.
• Fixed buffered senders requeueing only the first action after a request failure and asynchronous
  auto-flush cancelling its own indexing request.
• Fixed buffered senders ignoring the configured audience and omitting API version, credential
  scopes, transport, and policy options from schema lookup.
• Fixed document indexing rejecting caller-supplied error mappings instead of merging them with
  the default mapping, including when splitting oversized batches.
• Fixed resource deletes by name ignoring the requested match condition and caller-supplied ETag.
• Fixed field builders silently discarding additional supported field properties while preserving
  their existing defaults, selected field types, and hidden behavior.