{"id":6185,"date":"2025-11-18T08:05:30","date_gmt":"2025-11-18T16:05:30","guid":{"rendered":"https:\/\/devblogs.microsoft.com\/azure-sql\/?p=6185"},"modified":"2025-11-18T08:36:06","modified_gmt":"2025-11-18T16:36:06","slug":"public-preview-of-vector-indexing-in-azure-sql-db-azure-sql-mi-and-sql-database-in-microsoft-fabric","status":"publish","type":"post","link":"https:\/\/devblogs.microsoft.com\/azure-sql\/public-preview-of-vector-indexing-in-azure-sql-db-azure-sql-mi-and-sql-database-in-microsoft-fabric\/","title":{"rendered":"Public preview of vector indexing in Azure SQL DB, Azure SQL MI, and SQL database in Microsoft Fabric"},"content":{"rendered":"<p>We are happy to share that <strong>DiskANN vector indexing<\/strong> is now in <strong>public preview<\/strong> across Azure SQL Database, Azure SQL Managed Instance, and SQL database in Microsoft Fabric. DiskANN is a cutting-edge algorithm for efficient vector similarity search, making it ideal for powering <strong>recommendation systems<\/strong>, <strong>image and multimedia search<\/strong>, and advanced <strong>retrieval-augmented generation (RAG)<\/strong> solutions.<\/p>\n<pre class=\"prettyprint language-sql\"><code class=\"language-sql\">create vector index vec_idx on [dbo].[wikipedia_articles_embeddings]([title_vector]) \r\nwith (metric = 'cosine', type = 'diskann'); <\/code><\/pre>\n<p>Built for large-scale vector search, DiskANN integrates seamlessly with the Microsoft SQL query engine, enabling you to combine <strong>semantic search<\/strong> with the full power of T\u2011SQL, joins, filters, analytics, without moving data or adding a separate vector store. This is multimodal, AI-ready search <strong>inside the database engine<\/strong>.<\/p>\n<p>This release builds on the <strong>general availability of vector data type and vector functions announced earlier this summer<\/strong>, marking the next step toward embedding foundational AI capabilities directly in the database.<\/p>\n<p>Modern AI applications depend on fast approximate nearest neighbor (ANN) search over embeddings. Until now, teams often had to stitch together separate vector databases and ETL pipelines, adding complexity and cost. With DiskANN in Azure SQL and Fabric SQL:<\/p>\n<ul>\n<li><strong>Keep data where it lives: <\/strong>benefit from SQL\u2019s security, governance, and performance while adding high-performance vector search.<\/li>\n<li><strong>Compose rich queries: <\/strong>blend vector similarity with business filters and joins for multimodal experiences.<\/li>\n<li><strong>Scale confidently: <\/strong>DiskANN was designed for very large vector collections, balancing <strong>accuracy, query speed, and resource consumption<\/strong>.<\/li>\n<\/ul>\n<p><div class=\"alert alert-info\"><p class=\"alert-divider\"><i class=\"fabric-icon fabric-icon--Info\"><\/i><strong>Public Preview Deployment<\/strong><\/p>Public preview starts with Azure SQL Database and SQL database in Microsoft Fabric, with Azure SQL Managed Instance following shortly. Make sure to follow this blog to stay up-to-date with the latest and greatest news on this technology and more.<\/div><\/p>\n<p>For more details on vectors and vectors index, take a look also at this blog post: <a href=\"https:\/\/devblogs.microsoft.com\/azure-sql\/sql-server-2025-embraces-vectors-setting-the-foundation-for-empowering-your-data-with-ai\">SQL Server 2025 Embraces Vectors: setting the foundation for empowering your data with AI<\/a><\/p>\n<h2>Public Preview Notes<\/h2>\n<p>We\u2019ve received significant demand for vector indexing support, so we\u2019re releasing this feature early to unblock scenarios that can benefit from it. While the feature is available now, please note that there are some current limitations that may affect adoption in certain use cases. These limitations will be removed in the coming months: stay tuned for updates here.<\/p>\n<table style=\"border-collapse: collapse; width: 100%; border: none;\">\n<tbody>\n<tr style=\"background-color: unset;\">\n<td style=\"width: 5%; border: 0px;\"><\/td>\n<td style=\"width: 95%; border: 0px;\"><strong>Read-Only Tables\n<\/strong>When a table has a vector index, it becomes read-only. No data modifications are allowed while the vector index exists.\n<em>In Azure SQL Database and SQL Database in Microsoft Fabric, you can enable the <a href=\"https:\/\/learn.microsoft.com\/en-us\/sql\/t-sql\/statements\/create-vector-index-transact-sql#limitations\"><code>ALLOW_STALE_VECTOR_INDEX<\/code><\/a> database-scoped configuration to <code>ON<\/code>, which makes the table writable again.\u00a0<\/em><\/p>\n<p><strong>Primary Key Requirement\n<\/strong>The table must have a single-column, integer, primary key clustered index. <em>(Note: Embeddings and vectors should <a href=\"https:\/\/devblogs.microsoft.com\/azure-sql\/efficiently-and-elegantly-modeling-embeddings-in-azure-sql-and-sql-server\/\">ideally reside in their own table<\/a>)<\/em><\/p>\n<p><strong>Predicate Application\n<\/strong>Vector search is performed first, and <a href=\"https:\/\/learn.microsoft.com\/en-us\/sql\/t-sql\/functions\/vector-search-transact-sql#post-filter-only\">additional predicates are applied only after the most similar vectors are returned<\/a>.<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>We\u2019ll update this post as soon as these limitations are removed, in the meantime you can start to use vectors and vector index to simplify your AI solution, together with all the other enterprise-ready features in Azure SQL and SQL database in Fabric!<\/p>\n","protected":false},"excerpt":{"rendered":"<p>We are happy to share that DiskANN vector indexing is now in public preview across Azure SQL Database, Azure SQL Managed Instance, and SQL database in Microsoft Fabric. DiskANN is a cutting-edge algorithm for efficient vector similarity search, making it ideal for powering recommendation systems, image and multimedia search, and advanced retrieval-augmented generation (RAG) solutions. [&hellip;]<\/p>\n","protected":false},"author":24720,"featured_media":6198,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[1],"tags":[676,640,627,569,677],"class_list":["post-6185","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-azure-sql","tag-diskann","tag-embeddings","tag-rag","tag-vector","tag-vector-index"],"acf":[],"blog_post_summary":"<p>We are happy to share that DiskANN vector indexing is now in public preview across Azure SQL Database, Azure SQL Managed Instance, and SQL database in Microsoft Fabric. DiskANN is a cutting-edge algorithm for efficient vector similarity search, making it ideal for powering recommendation systems, image and multimedia search, and advanced retrieval-augmented generation (RAG) solutions. [&hellip;]<\/p>\n","_links":{"self":[{"href":"https:\/\/devblogs.microsoft.com\/azure-sql\/wp-json\/wp\/v2\/posts\/6185","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/devblogs.microsoft.com\/azure-sql\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/devblogs.microsoft.com\/azure-sql\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/devblogs.microsoft.com\/azure-sql\/wp-json\/wp\/v2\/users\/24720"}],"replies":[{"embeddable":true,"href":"https:\/\/devblogs.microsoft.com\/azure-sql\/wp-json\/wp\/v2\/comments?post=6185"}],"version-history":[{"count":0,"href":"https:\/\/devblogs.microsoft.com\/azure-sql\/wp-json\/wp\/v2\/posts\/6185\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/devblogs.microsoft.com\/azure-sql\/wp-json\/wp\/v2\/media\/6198"}],"wp:attachment":[{"href":"https:\/\/devblogs.microsoft.com\/azure-sql\/wp-json\/wp\/v2\/media?parent=6185"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/devblogs.microsoft.com\/azure-sql\/wp-json\/wp\/v2\/categories?post=6185"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/devblogs.microsoft.com\/azure-sql\/wp-json\/wp\/v2\/tags?post=6185"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}