With Elasticsearch Vector Database, Elastic's new serverless offering for vector workloads, Azure customers can build semantic search, retrieval-augmented generation (RAG), and agent-powered experiences. Azure Native Integrations make it easy to discover, deploy, and manage the resource through the Azure portal.

Built for AI, without managing a search cluster

Elasticsearch Vector Database brings key retrieval capabilities together, so you can focus on your application:

• Hybrid search and relevance: Combine keyword and vector retrieval, with metadata filters and reranking to refine the context returned to your application.
• Managed embeddings and inference: Use Elastic's semantic_text field and managed inference to generate embeddings without running separate machine learning infrastructure.
• Context for agents: Give agents access to relevant Elasticsearch data and tools through integrations such as Elastic MCP Server in Microsoft Foundry.

Built on Elastic's Search AI Lake architecture with Azure Blob Storage, Elastic Cloud Serverless manages the underlying infrastructure, so you don't have to provision or maintain clusters.

Discover and deploy directly from Azure

1. From the resource hub: Open Elastic Cloud (Elasticsearch) in the Azure portal. Select Create, then Elasticsearch Vector Database.
2. From Azure Marketplace: Open Elasticsearch - Vector Database to review the offering and begin creating your resource.

Discover Elasticsearch Vector Database alongside Elastic's search, observability, and security offerings.

You'll need deployment permissions, such as Owner or Contributor access, on the target Azure subscription. Follow the creation flow to select your subscription, resource group, and an available region, then review the configuration and create your resource.

Go from your resource to your first query

Your resource overview in the Azure portal brings together connection links and shortcuts to Elastic's tools:

• Add data & embed: Add documents and generate embeddings using semantic_text and Jina models.
• Open Agent Builder: Ask natural-language questions grounded in your data.
• Open in Dev Tools: Start with a prefilled semantic search example, then adapt it to your index and queries.

Start in Azure, then move directly into Elastic's tools to ingest, embed, and query your data.

Try Elastic's vector and full-text search quickstart, then use the retrieved context in an application powered by Azure OpenAI.

Keep management and billing connected to Azure

The Azure native integration connects more than the creation experience:

• Familiar management: Organize Elastic resources with Azure resource groups and tags, and view them alongside your other Azure resources.
• Connected access: Use Microsoft Entra ID single sign-on to access Elastic from Azure. Manage the Azure resource in the portal, then use Elastic's tools for data and search workflows.
• Consolidated billing: Marketplace purchases appear on your Azure billing statement. View Elastic usage and costs in the Elastic Cloud Console. Eligible purchases also count toward your Microsoft Azure Consumption Commitment (MACC), subject to offer eligibility and agreement terms.

Start building, and explore more on Azure

Start your next AI project with Elasticsearch - Vector Database in Azure Marketplace, or open the Elastic resource hub in the Azure portal. New users get a 7-day free trial!

Explore other Elastic offerings on Azure: Elasticsearch, Elastic Observability, and Elastic Security. Learn more about Elastic on Azure and how Azure Native Integrations connect partner services with Azure onboarding, management, identity, and billing.