From 3be95c80fa37895e4836b583237591e7d517ca50 Mon Sep 17 00:00:00 2001 From: Ben Verhees Date: Tue, 11 Nov 2025 23:14:56 +0100 Subject: [PATCH] docs: add vector search configuration for Azure AI Search --- pages/docs/configuration/dotenv.mdx | 1 + .../configuration/tools/azure_ai_search.mdx | 23 ++++++++++++++++--- 2 files changed, 21 insertions(+), 3 deletions(-) diff --git a/pages/docs/configuration/dotenv.mdx b/pages/docs/configuration/dotenv.mdx index 4b8f88950..296df1497 100644 --- a/pages/docs/configuration/dotenv.mdx +++ b/pages/docs/configuration/dotenv.mdx @@ -435,6 +435,7 @@ This plugin supports searching Azure AI Search for answers to your questions. Se ['AZURE_AI_SEARCH_SEARCH_OPTION_QUERY_TYPE', 'string', 'The query type for Azure AI Search.','AZURE_AI_SEARCH_SEARCH_OPTION_QUERY_TYPE='], ['AZURE_AI_SEARCH_SEARCH_OPTION_TOP', 'number', 'The top count for Azure AI Search.','AZURE_AI_SEARCH_SEARCH_OPTION_TOP='], ['AZURE_AI_SEARCH_SEARCH_OPTION_SELECT', 'string', 'The select fields for Azure AI Search.','AZURE_AI_SEARCH_SEARCH_OPTION_SELECT='], + ['AZURE_AI_SEARCH_VECTOR_FIELDS', 'string', 'The names of the vector fields in your Azure AI Search index to use for vector search queries. Multiple fields can be specified as comma-separated values (e.g., \'field1,field2\'). Leave empty if you don\'t use vector search.','AZURE_AI_SEARCH_VECTOR_FIELDS='], ]} /> diff --git a/pages/docs/configuration/tools/azure_ai_search.mdx b/pages/docs/configuration/tools/azure_ai_search.mdx index e9a557d54..41fad623e 100644 --- a/pages/docs/configuration/tools/azure_ai_search.mdx +++ b/pages/docs/configuration/tools/azure_ai_search.mdx @@ -103,7 +103,11 @@ Now select the free option or select your preferred option (may incur charges). ![image](/images/azure-ai-search/librechat_settings.png) -**2.** Fill in the Endpoint, Index Name, and API Key, and click on `Save`. +**2.** Fill in the Endpoint, Index Name, and API Key. + +**3.** (Optional) If you're using integrated vectorization, specify the Azure AI Search Vector Fields field with the comma-separated names of your vector fields (e.g., `vectorField1,vectorField2`). See [AZURE_AI_SEARCH_VECTOR_FIELDS](#azure_ai_search_vector_fields) for more details on vector search configuration. + +**4.** Click on `Save` to save your configuration. # Conclusion @@ -121,15 +125,16 @@ For details on each parameter, please refer to the following document: **[https://learn.microsoft.com/en-us/rest/api/searchservice/search-documents](https://learn.microsoft.com/en-us/rest/api/searchservice/search-documents)** ```env -AZURE_AI_SEARCH_API_VERSION=2023-10-01-Preview +AZURE_AI_SEARCH_API_VERSION=2025-09-01 AZURE_AI_SEARCH_SEARCH_OPTION_QUERY_TYPE=simple AZURE_AI_SEARCH_SEARCH_OPTION_TOP=3 AZURE_AI_SEARCH_SEARCH_OPTION_SELECT=field1, field2, field3 +AZURE_AI_SEARCH_VECTOR_FIELDS=vectorField1,vectorField2 ``` #### AZURE_AI_SEARCH_API_VERSION -Specify the version of the search API. When using new features such as semantic search or vector search, you may need to specify the preview version. The default value is `2023-11-1`. +Specify the version of the search API. When using new features such as semantic search or vector search, you may need to specify the preview version. The default value is `2025-09-01`. #### AZURE_AI_SEARCH_SEARCH_OPTION_QUERY_TYPE @@ -142,3 +147,15 @@ Specify the number of items to search for. The default value is 5. #### AZURE_AI_SEARCH_SEARCH_OPTION_SELECT Specify the fields of the index to be retrieved, separated by commas. Please note that these are not the fields to be searched. + +#### AZURE_AI_SEARCH_VECTOR_FIELDS + +Specify the names of the vector fields in your Azure AI Search index to use for vector search queries. Multiple fields can be specified as comma-separated values (e.g., `field1,field2`). Leave empty if you don't use vector search. + +When configured, this enables vector search capabilities in the Azure AI Search tool. The tool will accept a `vectorQueryText` parameter that will be vectorized and used to perform semantic similarity searches across the specified vector fields. You can use both keyword search (`query`) and vector search (`vectorQueryText`) together for hybrid search scenarios. + +**Prerequisite:** Your Azure AI Search index must have [integrated vectorization](https://learn.microsoft.com/en-us/azure/search/vector-search-integrated-vectorization#using-integrated-vectorization-in-queries) configured with vectorizers assigned to your vector fields. The field names specified in `AZURE_AI_SEARCH_VECTOR_FIELDS` must match the actual vector field names in your index that have vectorizers assigned. + +For more information on setting up vector search in Azure AI Search, see: +- [Azure AI Search vector search overview](https://learn.microsoft.com/en-us/azure/search/vector-search-overview) +- [Integrated vectorization in queries](https://learn.microsoft.com/en-us/azure/search/vector-search-integrated-vectorization#using-integrated-vectorization-in-queries)