From fc87d079c60ed9fcefeaea13361cb0c788ba0347 Mon Sep 17 00:00:00 2001 From: Vlada Chirmicci Date: Fri, 17 Jul 2026 17:56:59 +0100 Subject: [PATCH 01/13] Adding Vector Database project type docs --- deploy-manage/cloud-organization/billing.md | 1 + .../serverless-project-billing-dimensions.md | 1 + .../vector-database-billing-dimensions.md | 99 ++++++++++++++++ deploy-manage/toc.yml | 1 + solutions/index.md | 2 +- solutions/toc.yml | 3 + solutions/vector-database.md | 76 ++++++++++++ solutions/vector-database/get-started.md | 111 ++++++++++++++++++ 8 files changed, 293 insertions(+), 1 deletion(-) create mode 100644 deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md create mode 100644 solutions/vector-database.md create mode 100644 solutions/vector-database/get-started.md diff --git a/deploy-manage/cloud-organization/billing.md b/deploy-manage/cloud-organization/billing.md index 81d72a5b0f..f62660b497 100644 --- a/deploy-manage/cloud-organization/billing.md +++ b/deploy-manage/cloud-organization/billing.md @@ -23,6 +23,7 @@ Elastic charges a recurring fee for using our offerings on {{ecloud}}. In this s * [](/deploy-manage/cloud-organization/billing/cloud-hosted-deployment-billing-dimensions.md) * [](/deploy-manage/cloud-organization/billing/serverless-project-billing-dimensions.md) * [](/deploy-manage/cloud-organization/billing/elasticsearch-billing-dimensions.md) + * [](/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md) * [](/deploy-manage/cloud-organization/billing/elastic-observability-billing-dimensions.md) * [](/deploy-manage/cloud-organization/billing/security-billing-dimensions.md) diff --git a/deploy-manage/cloud-organization/billing/serverless-project-billing-dimensions.md b/deploy-manage/cloud-organization/billing/serverless-project-billing-dimensions.md index 5b987ff4de..c8da09260e 100644 --- a/deploy-manage/cloud-organization/billing/serverless-project-billing-dimensions.md +++ b/deploy-manage/cloud-organization/billing/serverless-project-billing-dimensions.md @@ -22,6 +22,7 @@ Specific prices can be found in the [Cloud Pricing Table](https://cloud.elastic. To learn about billing dimensions for specific offerings, refer to: * [](elasticsearch-billing-dimensions.md) +* [](vector-database-billing-dimensions.md) * [](elastic-observability-billing-dimensions.md) * [](security-billing-dimensions.md) diff --git a/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md b/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md new file mode 100644 index 0000000000..67772647b9 --- /dev/null +++ b/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md @@ -0,0 +1,99 @@ +--- +navigation_title: Vector Database projects +applies_to: + serverless: preview +products: + - id: cloud-serverless + - id: elasticsearch +description: >- + Learn how costs for Elasticsearch Vector Database Serverless projects are + calculated across storage, searchable capacity, indexing, and infrastructure. +--- + +# {{es}} Vector Database billing dimensions [vector-database-billing] + +{{es}} Vector Database projects on {{serverless-full}} use a different billing model from {{es-serverless}} projects. Instead of virtual compute units (VCUs), charges are based on storage, searchable capacity, indexing volume, and project infrastructure time. + +For current rates, refer to the [Cloud Pricing Table](https://cloud.elastic.co/cloud-pricing-table?productType=serverless) or the [{{serverless-full}} pricing](https://www.elastic.co/pricing/serverless) pages. + + + +## Boosted and unboosted indices [vector-database-billing-boosted-unboosted] + +Indices in a Vector Database project can be **boosted** or **unboosted**. These states describe searchability only: + +| Index state | Meaning | +| --- | --- | +| **Boosted** | The index is searchable. | +| **Unboosted** | The index is not searchable. | + +Do not treat boosted vs unboosted as billing tiers. Storage charges still apply to durable data in either state. How the invoice **Boosted** line item is calculated in the initial release is covered under [Boosted (searchable capacity)](#vector-database-billing-boosted). + +::::{warning} +* **Unboosted indices are not searchable.** +* **In the initial release, billing is still computed as if all data is boosted.** Changing an index to unboosted does not reduce the Boosted line item. There is no archive or unboosted discount yet. +:::: + +## Billing dimensions [vector-database-billing-dimensions] + +Your invoice includes these line items: + +| Line item | What you pay for | How it is measured | +| --- | --- | --- | +| **Storage** | Durable data stored in the project | GB-month of stored data | +| **Boosted** | Searchable capacity for your data | GB-month of boosted data, multiplied by Search Power | +| **Indexing** | Write volume into the project | GB of billable indexing over the billing period | +| **Infrastructure** | Keeping the project available | Hours the project is running | + +### Storage [vector-database-billing-storage] + +**Storage** covers the durable footprint of your indices, measured in GB-month (decimal gigabytes, prorated over time). + +You are charged for data as long as it remains stored in the project, whether or not it is searchable. + +### Boosted (searchable capacity) [vector-database-billing-boosted] + +The invoice line item named **Boosted** covers the cost of searchable capacity. Charges use GB-month of data priced as boosted, scaled by your project's [Search Power](/deploy-manage/deploy/elastic-cloud/project-settings.md#elasticsearch-manage-project-search-power-settings) (SP): + +```text +Boosted charge ∝ Boosted_GB × (SP_used / 100) +``` + +Higher Search Power increases searchable performance and increases the Boosted line item. If Search Power changes during the billing period, or if the project autoscales within a configured SP range, Elastic bills using the time-weighted Search Power actually used. + +In the initial release, `Boosted_GB` matches total durable storage for pricing. See [Boosted and unboosted indices](#vector-database-billing-boosted-unboosted). + +### Indexing [vector-database-billing-indexing] + +**Indexing** covers billable write volume into the project, measured in GB for the billing period. This follows the same platform definition of billable indexing bytes used for other {{serverless-short}} offerings. + +### Infrastructure [vector-database-billing-infrastructure] + +**Infrastructure** is a project availability fee billed by the hour for the time your project is running. + +Marketing may describe this as a monthly amount (for example, a full month of continuous availability). On the invoice, that amount is converted to an hourly rate and charged for the hours the project emits a heartbeat. If the project runs for only part of the month, you pay only for those hours. + +## How this differs from {{es-serverless}} [vector-database-billing-vs-elasticsearch] + +| | {{es}} Vector Database | {{es-serverless}} | +| --- | --- | --- | +| Primary model | Storage, boosted capacity, indexing GB, and infrastructure hours | VCUs (search, ingest, ML) plus storage | +| Search performance lever | Search Power multiplies the Boosted line | Search Power affects Search VCU baseline and scale | +| Idle project | Infrastructure accrues while the project exists; storage and boosted continue for stored data | Search VCUs keep a reduced baseline while data remains searchable | + +For {{es-serverless}} billing details, see [](elasticsearch-billing-dimensions.md). + +## Understand your bill [vector-database-billing-explain] + +Use this mapping to relate each invoice line item to the underlying meter: + +| Invoice line item | Meter | What drives the charge | +| --- | --- | --- | +| **Storage** | `Storage_GB` (GB-month) | Durable bytes stored over time for all indices (boosted and unboosted) | +| **Boosted** | `Boosted_GB` (GB-month) × (`SP_used` / 100) | Data priced as searchable capacity, scaled by the Search Power actually used over time | +| **Indexing** | `Indexing_GB` (GB) | Billable write volume during the billing period | +| **Infrastructure** | Hours running | Hours the project is available (heartbeat), at the published hourly infrastructure rate | + +In the initial release, `Boosted_GB` equals `Storage_GB` for billing. Search Power changes and autoscaling within an SP range are time-weighted into `SP_used`. + +Shared {{serverless-short}} add-ons such as [data out](serverless-project-billing-dimensions.md#general-serverless-billing-data-out) and [support](serverless-project-billing-dimensions.md#general-serverless-billing-support) may also appear on your bill. See [](serverless-project-billing-dimensions.md). diff --git a/deploy-manage/toc.yml b/deploy-manage/toc.yml index 3a8e244584..bd921e93e7 100644 --- a/deploy-manage/toc.yml +++ b/deploy-manage/toc.yml @@ -809,6 +809,7 @@ toc: - file: cloud-organization/billing/serverless-project-billing-dimensions.md children: - file: cloud-organization/billing/elasticsearch-billing-dimensions.md + - file: cloud-organization/billing/vector-database-billing-dimensions.md - file: cloud-organization/billing/elastic-observability-billing-dimensions.md - file: cloud-organization/billing/security-billing-dimensions.md - file: cloud-organization/billing/billing-models.md diff --git a/solutions/index.md b/solutions/index.md index 5e56e7b3ed..fd08e6ea9e 100644 --- a/solutions/index.md +++ b/solutions/index.md @@ -23,7 +23,7 @@ Elastic helps you build applications for three main use cases: search, observabi | Your use case | What to use | Description | | --- | --- | --- | -| Building search-powered applications | 1. [Core search capabilities](/solutions/search.md)

2. [Elasticsearch solution](/solutions/elasticsearch-solution-project.md) | 1. Core {{es}} search features available across all deployment types, solutions, and project types

2. Additional UI tools that complement the core search features | +| Building search-powered applications | 1. [Core search capabilities](/solutions/search.md)

2. [Elasticsearch solution](/solutions/elasticsearch-solution-project.md)

3. [{{es}} Vector Database](/solutions/vector-database.md) {applies_to}`serverless: preview` | 1. Core {{es}} search features available across all deployment types, solutions, and project types

2. Additional UI tools that complement the core search features

3. Dedicated {{serverless-full}} project type for AI-powered retrieval (RAG, recommendations, semantic and hybrid search) with vector-tuned defaults | | Monitoring applications or infrastructure | [Observability solution](/solutions/observability.md) | Monitor and troubleshoot with logs, metrics, and traces | | Protecting against threats | [Security solution](/solutions/security.md) | Detect and respond to security threats | diff --git a/solutions/toc.yml b/solutions/toc.yml index 2c24b54080..13213f033e 100644 --- a/solutions/toc.yml +++ b/solutions/toc.yml @@ -85,6 +85,9 @@ toc: - file: elasticsearch-solution-project/search-applications/search-application-api.md - file: elasticsearch-solution-project/search-applications/search-application-security.md - file: elasticsearch-solution-project/search-applications/search-application-client.md + - file: vector-database.md + children: + - file: vector-database/get-started.md - file: observability.md children: - file: observability/get-started.md diff --git a/solutions/vector-database.md b/solutions/vector-database.md new file mode 100644 index 0000000000..03c890abcb --- /dev/null +++ b/solutions/vector-database.md @@ -0,0 +1,76 @@ +--- +navigation_title: Vector Database project +applies_to: + serverless: preview +description: >- + The Elasticsearch Vector Database project type on Elastic Cloud Serverless is + built for AI-powered retrieval workloads such as RAG, recommendations, and semantic search. +products: + - id: elasticsearch + - id: cloud-serverless +--- + +# {{es}} Vector Database project + +The {{es}} Vector Database {{serverless-short}} project type provides optimized defaults for vector search workloads. Use it when you are building AI-powered retrieval, such as a chatbot, RAG pipeline, or recommendation engine, where embeddings and similarity search are central. Typical use cases include RAG, semantic retrieval, recommendations, and hybrid search. + +[Vector search](/solutions/search/vector.md) works the same way as in other {{es}} projects. The difference is that the default configuration of the Vector Database project is tuned for vector-first workloads. + +## What you get + +The {{es}} Vector Database project type is designed for load-once, query-often vector workloads that your application calls over the API. Compared with a general-purpose {{es-serverless}} project, it prioritizes vector-friendly index defaults suited to embedding storage and similarity search. + +### Vector index mode by default + +Indices in a Vector Database project use the [`index.mode: vectordb_document`](elasticsearch://reference/elasticsearch/mapping-reference/dense-vector.md#dense-vector-vectordb-document-mode) vector index mode automatically. It applies storage, indexing, and merge defaults tuned for similarity search on dense vectors, so you get efficient embedding storage and approximate [kNN](/solutions/search/vector/knn.md) search without configuring each setting yourself. + +:::{tip} +On other [deployment types](/deploy-manage/deploy.md), you can set the vector index mode explicitly when you create an index. +::: + +### Built for vector query patterns + +The Vector Database project type favors workloads where you ingest and embed data, then serve similarity or hybrid queries repeatedly. Aggressive segment merging improves recall and query speed for relatively stable corpora, which is a common pattern for knowledge bases, product catalogs with semantic search, and RAG document stores. + + + +### Multi-tenant partitioning with slices + +For multi-tenant or multi-customer vector apps, you can use slices to partition an index so each tenant’s vectors are indexed and searched in isolation. Slice mode is opt-in and is not enabled automatically on Vector Database projects. + + + +### Usage-based {{serverless-short}} operations + +Similar to other {{serverless-full}} projects, Elastic manages the infrastructure, scaling, and upgrades. You create a project, get an endpoint, and start indexing and querying without sizing nodes for vector RAM yourself. + +Billing uses storage, searchable capacity, indexing volume, and infrastructure hours rather than the VCU model used by {{es-serverless}}. See [{{es}} Vector Database billing dimensions](/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md). + +## When to use this project type + +Choose {{es}} Vector Database when you are building retrieval for AI features: a chatbot, RAG pipeline, recommendation engine, or similar workloads where your application consumes results. This provides out of the box vector-tuned defaults. + +Choose the [{{es}} project type](/solutions/elasticsearch-solution-project.md) when you are building a search application people interact with directly (a search bar, catalog, or knowledge base), or when you need broader search application tooling in {{kib}} such as [Playground](/solutions/elasticsearch-solution-project/playground.md), [Query Rules UI](/solutions/elasticsearch-solution-project/query-rules-ui.md), or [Agent Builder](/explore-analyze/ai-features/elastic-agent-builder.md). + +| Use case | Fit | Why | +| --- | --- | --- | +| [RAG and question answering](/solutions/search/vector/vector-search-use-cases.md#rag-and-question-answering-on-your-own-data) | Strong | Retrieve passages from documents, wikis, tickets, or knowledge bases and pass them to an LLM for assistants, support bots, and cited answers | +| [Discovery and recommendations](/solutions/search/vector/vector-search-use-cases.md#discovery-and-recommendations) | Strong | Find related products, articles, or other items by similarity when keywords alone are not enough | +| [Multimodal search](/solutions/search/vector/vector-search-use-cases.md#multimodal-search) | Strong | Search across images, audio, video, or text with embeddings from a multimodal model | +| [Duplicate detection, fraud, and anomaly detection](/solutions/search/vector/vector-search-use-cases.md#duplicate-detection-fraud-and-anomaly-detection) | Strong | Compare embeddings to find near-duplicates, suspicious matches, or unusual patterns at scale | +| [Long-term memory for LLMs](/solutions/search/vector/vector-search-use-cases.md#long-term-memory-for-llms) | Strong | Store facts, chat turns, or summaries so an assistant can retrieve relevant past context | +| Full-text or keyword search without vectors | Prefer the {{es}} project | The {{es}} project is built for lexical search, filters, and analytics on document-centric data | +| Log, event, or other time series search | Prefer {{es}} project | Choose the {{es}} project for continuously ingested logs and events, where general-purpose defaults fit write-heavy, frequently updated data | + +## Get started + +Ready to try the Vector Database project type? Follow [Get started](/solutions/vector-database/get-started.md) to create a project, ingest embeddings, and run your first searches. + +## Related pages + +* [Vector search in {{es}}](/solutions/search/vector.md) +* [Vector search use cases](/solutions/search/vector/vector-search-use-cases.md) +* [Semantic search](/solutions/search/semantic-search.md) +* [{{es}} solution overview](/solutions/elasticsearch-solution-project.md) +* [Search use case documentation](/solutions/search.md) +* [{{es-serverless}} API documentation]({{es-serverless-apis}}) diff --git a/solutions/vector-database/get-started.md b/solutions/vector-database/get-started.md new file mode 100644 index 0000000000..e9f46e1b23 --- /dev/null +++ b/solutions/vector-database/get-started.md @@ -0,0 +1,111 @@ +--- +navigation_title: Get started +description: >- + Create an Elasticsearch Vector Database project on Elastic Cloud Serverless, + ingest embeddings, and run your first vector or semantic searches. +applies_to: + serverless: preview +products: + - id: elasticsearch + - id: cloud-serverless +--- + +# Get started with the {{es}} Vector Database project type + +The {{es}} Vector Database project type on {{serverless-full}} is built for AI-powered vector retrieval. Use this guide to create a project, add embeddings, and run similarity or semantic searches from your application. + +:::{note} +Not sure whether this project type is right for you? Refer to [When to use this project type](/solutions/vector-database.md#when-to-use-this-project-type). + +If you're looking for an introduction to the {{stack}} or the {{es}} product, go to [](/get-started/index.md) or [](/manage-data/data-store.md). +::: + +::::::{stepper} +:::::{step} Create an {{es}} Vector Database {{serverless-short}} project + +There are two options to create serverless projects: + +* If you're a new user, [sign up for a free 14-day trial](https://cloud.elastic.co/serverless-registration). For more information about {{ecloud}} trials, check out [Trial information](/deploy-manage/deploy/elastic-cloud/create-an-organization.md#general-sign-up-trial-what-is-included-in-my-trial). +* If you're an existing customer, [log in to {{ecloud}}](https://cloud.elastic.co/login) and do the following: + 1. Select **Create project** from the **Serverless projects** panel. + 2. Select **Next** from the **Vector Database** panel. + 3. Name your project. + 4. Select a cloud provider and region. For available regions, refer to [Regions](/deploy-manage/deploy/elastic-cloud/regions.md). + 5. Select **Create project**. It takes a few minutes to create your project. + 6. When the project is ready, select **Continue** to open it (you might need to log in to {{ecloud}} again). + +:::{note} +You need the `admin` predefined role or an equivalent custom role to create projects. For more information, refer to [User roles and privileges](/deploy-manage/users-roles/cloud-organization/user-roles.md). +::: + +After you've created your project, note the {{es}} endpoint and API key from the project connection details. You'll use these to index data and run searches. New indices in this project type use [vector index mode](elasticsearch://reference/elasticsearch/mapping-reference/dense-vector.md#dense-vector-vectordb-document-mode) automatically. +::::: + +:::::{step} (Optional) Follow the in-product setup guides + +When you create a new Vector Database project, the **Set up your Elasticsearch Vector Database** page includes two guided paths you can follow. Each path walks you through ingest and then search examples, with sample scripts you can run in a client of your choosing or run the examples directly in the Console. + +Choose one: + +| Guide | When to use | What you do | +| --- | --- | --- | +| Generate embeddings from your content | You want {{es}} to create embeddings for you | Ingest content into a `semantic_text` field, then run a semantic or hybrid query | +| Store your existing embeddings | You already have vectors from your own model | Index pre-generated embeddings into a `dense_vector` field, then run a semantic or hybrid query | + +You can also skip the setup guide and continue with the steps below. +::::: + +:::::{step} Ingest your data +Use the approach that matches how you create embeddings. + +::::{dropdown} Generate embeddings from your content +Create an index with a [`semantic_text`](elasticsearch://reference/elasticsearch/mapping-reference/semantic-text.md) field, then ingest documents so {{es}} generates embeddings automatically. The default model is [`jina-embeddings-v3`](/explore-analyze/machine-learning/nlp/ml-nlp-jina.md#jina-embeddings-v3), integrated through the {{es}} {{infer}} pipeline. + +:::: + +::::{dropdown} Store your existing embeddings +Create an index with a [`dense_vector`](elasticsearch://reference/elasticsearch/mapping-reference/dense-vector.md) field, and any other text or metadata fields you might need, then index your pre-generated vectors. + +To walk through indexing sample embeddings and running a kNN search, follow [Bring your own dense vectors to {{es}}](/solutions/search/vector/bring-own-vectors.md). +:::: + +Use a language [client](/reference/elasticsearch-clients/index.md) or the [Bulk]({{es-serverless-apis}}operation/operation-bulk) API to load data. If you're not ready to use your own content, follow the [semantic search quickstart](/solutions/search/get-started/semantic-search.md). +::::: + +:::::{step} Search your data +Use the approach that matches how you ingested embeddings. + +::::{dropdown} Search embeddings generated from your content +Query with natural language. {{es}} embeds the query and returns the most relevant matches. You can also run a hybrid query that combines semantic and lexical ranking. + +To walk through hybrid search with `semantic_text`, follow [Hybrid search with `semantic_text`](/solutions/search/hybrid-semantic-text.md). For query and retrieval options, see [Search and retrieve `semantic_text` fields](elasticsearch://reference/elasticsearch/mapping-reference/semantic-text-search-retrieval.md). +:::: + +::::{dropdown} Search your existing embeddings +Run a kNN search against your `dense_vector` field, or combine kNN with lexical search for hybrid retrieval. + +Queries need to be converted into vectors the same way your documents were. Specify the model you used to generate your embeddings so {{es}} can apply it to your queries too. + +To walk through kNN search, follow [kNN search in {{es}}](/solutions/search/vector/knn.md). For combining vector and lexical ranking, see [Hybrid search](/solutions/search/hybrid-search.md). +:::: + +For more query options, see [Querying for search](/solutions/search/querying-for-search.md) and the [{{es-serverless}} API documentation]({{es-serverless-apis}}). +::::: + +:::::{step} Tune for your workload + +After you have a baseline that returns relevant results: + +* Review [vector search](/solutions/search/vector.md) concepts (field types, quantization, similarity) +* Adjust relevance with [ranking and reranking](/solutions/search/ranking.md) +* For RAG applications, see [RAG](/solutions/search/rag.md) +* Review project [Search Power](/deploy-manage/deploy/elastic-cloud/project-settings.md#elasticsearch-manage-project-search-power-settings) and [billing dimensions](/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md) for latency versus cost +::::: +:::::: + +## Next steps + +* [Vector search in {{es}}](/solutions/search/vector.md): Concepts, field types, and query options +* [Vector search use cases](/solutions/search/vector/vector-search-use-cases.md): RAG, recommendations, multimodal search, and more +* [{{es}} Vector Database overview](/solutions/vector-database.md): When to choose this project type +* [{{es}} solution](/solutions/elasticsearch-solution-project.md): UI tools for broader search application development From 42cb07909b2907833fc07abce3291563bc8118fc Mon Sep 17 00:00:00 2001 From: Vlada Chirmicci Date: Mon, 20 Jul 2026 14:18:24 +0100 Subject: [PATCH 02/13] Make a few more edits --- .../vector-database-billing-dimensions.md | 87 +++++-------------- solutions/vector-database.md | 37 +++++--- solutions/vector-database/get-started.md | 23 ++--- 3 files changed, 55 insertions(+), 92 deletions(-) diff --git a/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md b/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md index 67772647b9..526481e6fe 100644 --- a/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md +++ b/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md @@ -7,93 +7,52 @@ products: - id: elasticsearch description: >- Learn how costs for Elasticsearch Vector Database Serverless projects are - calculated across storage, searchable capacity, indexing, and infrastructure. + calculated across storage, search, and indexing. --- # {{es}} Vector Database billing dimensions [vector-database-billing] -{{es}} Vector Database projects on {{serverless-full}} use a different billing model from {{es-serverless}} projects. Instead of virtual compute units (VCUs), charges are based on storage, searchable capacity, indexing volume, and project infrastructure time. +{{es}} Vector Database projects on {{serverless-full}} are priced based on consumption across storage, search, and indexing. -For current rates, refer to the [Cloud Pricing Table](https://cloud.elastic.co/cloud-pricing-table?productType=serverless) or the [{{serverless-full}} pricing](https://www.elastic.co/pricing/serverless) pages. - - - -## Boosted and unboosted indices [vector-database-billing-boosted-unboosted] - -Indices in a Vector Database project can be **boosted** or **unboosted**. These states describe searchability only: - -| Index state | Meaning | -| --- | --- | -| **Boosted** | The index is searchable. | -| **Unboosted** | The index is not searchable. | - -Do not treat boosted vs unboosted as billing tiers. Storage charges still apply to durable data in either state. How the invoice **Boosted** line item is calculated in the initial release is covered under [Boosted (searchable capacity)](#vector-database-billing-boosted). +Your monthly bill is calculated based on these components: -::::{warning} -* **Unboosted indices are not searchable.** -* **In the initial release, billing is still computed as if all data is boosted.** Changing an index to unboosted does not reduce the Boosted line item. There is no archive or unboosted discount yet. -:::: +* **Storage** — Measured by the total amount of data stored in your project, in GB. +* **Search** — Measured by the resources allocated to keep your data searchable. Allocation is controlled by your project's [Search Power](/deploy-manage/deploy/elastic-cloud/project-settings.md#elasticsearch-manage-project-search-power-settings) setting. +* **Indexing** — Measured by the volume of data you index into your project over the course of a month, in GB. -## Billing dimensions [vector-database-billing-dimensions] -Your invoice includes these line items: - -| Line item | What you pay for | How it is measured | -| --- | --- | --- | -| **Storage** | Durable data stored in the project | GB-month of stored data | -| **Boosted** | Searchable capacity for your data | GB-month of boosted data, multiplied by Search Power | -| **Indexing** | Write volume into the project | GB of billable indexing over the billing period | -| **Infrastructure** | Keeping the project available | Hours the project is running | - -### Storage [vector-database-billing-storage] - -**Storage** covers the durable footprint of your indices, measured in GB-month (decimal gigabytes, prorated over time). - -You are charged for data as long as it remains stored in the project, whether or not it is searchable. - -### Boosted (searchable capacity) [vector-database-billing-boosted] +For current rates, refer to the [Cloud Pricing Table](https://cloud.elastic.co/cloud-pricing-table?productType=serverless) or the [{{serverless-full}} pricing](https://www.elastic.co/pricing/serverless) pages. -The invoice line item named **Boosted** covers the cost of searchable capacity. Charges use GB-month of data priced as boosted, scaled by your project's [Search Power](/deploy-manage/deploy/elastic-cloud/project-settings.md#elasticsearch-manage-project-search-power-settings) (SP): + -```text -Boosted charge ∝ Boosted_GB × (SP_used / 100) -``` +## Storage [vector-database-billing-storage] -Higher Search Power increases searchable performance and increases the Boosted line item. If Search Power changes during the billing period, or if the project autoscales within a configured SP range, Elastic bills using the time-weighted Search Power actually used. +You are charged per GB of data stored in the project. Storage charges apply for as long as the data remains in the project. -In the initial release, `Boosted_GB` matches total durable storage for pricing. See [Boosted and unboosted indices](#vector-database-billing-boosted-unboosted). +## Search [vector-database-billing-search] -### Indexing [vector-database-billing-indexing] +Search charges cover the resources used to keep your data searchable. All stored data is boosted by default, which means it is searchable and contributes to Search charges. -**Indexing** covers billable write volume into the project, measured in GB for the billing period. This follows the same platform definition of billable indexing bytes used for other {{serverless-short}} offerings. +[Search Power](/deploy-manage/deploy/elastic-cloud/project-settings.md#elasticsearch-manage-project-search-power-settings) controls how many baseline resources are kept ready for your data and the maximum resources the system can allocate under search load. Increasing Search Power improves query performance and increases search charges. Decreasing Search Power reduces provisioned resources and cost, with more variable query latency under load. -### Infrastructure [vector-database-billing-infrastructure] +## Indexing [vector-database-billing-indexing] -**Infrastructure** is a project availability fee billed by the hour for the time your project is running. +Indexing charges are based on the volume of data written to your project, measured in GB for the billing period. This uses the same definition of billable indexing bytes as other {{serverless-short}} offerings. -Marketing may describe this as a monthly amount (for example, a full month of continuous availability). On the invoice, that amount is converted to an hourly rate and charged for the hours the project emits a heartbeat. If the project runs for only part of the month, you pay only for those hours. +## Managing Vector Database costs [vector-database-billing-managing-costs] -## How this differs from {{es-serverless}} [vector-database-billing-vs-elasticsearch] +Vector Database costs follow your storage footprint, search resource allocation, and indexing volume. To balance performance with spend, adjust the controls described in this section. -| | {{es}} Vector Database | {{es-serverless}} | -| --- | --- | --- | -| Primary model | Storage, boosted capacity, indexing GB, and infrastructure hours | VCUs (search, ingest, ML) plus storage | -| Search performance lever | Search Power multiplies the Boosted line | Search Power affects Search VCU baseline and scale | -| Idle project | Infrastructure accrues while the project exists; storage and boosted continue for stored data | Search VCUs keep a reduced baseline while data remains searchable | +### Search Power setting [vector-database-billing-search-power-setting] -For {{es-serverless}} billing details, see [](elasticsearch-billing-dimensions.md). +[Search Power](/deploy-manage/deploy/elastic-cloud/project-settings.md#elasticsearch-manage-project-search-power-settings) controls search performance and the resources allocated to boosted data. Increase Search Power when you need more consistent latency and throughput. Decrease it when you want to reduce Search charges and can accept more variable latency. -## Understand your bill [vector-database-billing-explain] +### Inactivity API [vector-database-billing-inactivity-api] -Use this mapping to relate each invoice line item to the underlying meter: +Boosted versus unboosted is not a customer-facing index setting. All storage is boosted by default. To reduce Search charges for data you do not need to keep searchable, use the Inactivity API. -| Invoice line item | Meter | What drives the charge | -| --- | --- | --- | -| **Storage** | `Storage_GB` (GB-month) | Durable bytes stored over time for all indices (boosted and unboosted) | -| **Boosted** | `Boosted_GB` (GB-month) × (`SP_used` / 100) | Data priced as searchable capacity, scaled by the Search Power actually used over time | -| **Indexing** | `Indexing_GB` (GB) | Billable write volume during the billing period | -| **Infrastructure** | Hours running | Hours the project is available (heartbeat), at the published hourly infrastructure rate | + -In the initial release, `Boosted_GB` equals `Storage_GB` for billing. Search Power changes and autoscaling within an SP range are time-weighted into `SP_used`. +## Related billing dimensions [vector-database-billing-related] Shared {{serverless-short}} add-ons such as [data out](serverless-project-billing-dimensions.md#general-serverless-billing-data-out) and [support](serverless-project-billing-dimensions.md#general-serverless-billing-support) may also appear on your bill. See [](serverless-project-billing-dimensions.md). diff --git a/solutions/vector-database.md b/solutions/vector-database.md index 03c890abcb..cfca7dbbe5 100644 --- a/solutions/vector-database.md +++ b/solutions/vector-database.md @@ -4,25 +4,28 @@ applies_to: serverless: preview description: >- The Elasticsearch Vector Database project type on Elastic Cloud Serverless is - built for AI-powered retrieval workloads such as RAG, recommendations, and semantic search. + optimized for vector workloads, with vector-tuned defaults, hardware profile, + inference access, and pricing. It supports semantic and hybrid search. products: - id: elasticsearch - id: cloud-serverless --- -# {{es}} Vector Database project +# {{es}} Vector Database project overview -The {{es}} Vector Database {{serverless-short}} project type provides optimized defaults for vector search workloads. Use it when you are building AI-powered retrieval, such as a chatbot, RAG pipeline, or recommendation engine, where embeddings and similarity search are central. Typical use cases include RAG, semantic retrieval, recommendations, and hybrid search. +The {{es}} Vector Database {{serverless-short}} project type is optimized for vector workloads. Compared with the general-purpose [{{es}} project type](/solutions/elasticsearch-solution-project.md), it uses a vector-tuned default configuration, a hardware profile suited to embeddings, streamlined access to {{infer}}, and a pricing model built for vector storage and search. -[Vector search](/solutions/search/vector.md) works the same way as in other {{es}} projects. The difference is that the default configuration of the Vector Database project is tuned for vector-first workloads. +[Vector search](/solutions/search/vector.md) uses the same query APIs in both project types. + +Use it when embeddings and similarity search are central to your application, for example RAG, recommendations, semantic retrieval, or multimodal search. You can run [semantic](/solutions/search/semantic-search.md) and [hybrid](/solutions/search/hybrid-search.md) search in the same project. ## What you get -The {{es}} Vector Database project type is designed for load-once, query-often vector workloads that your application calls over the API. Compared with a general-purpose {{es-serverless}} project, it prioritizes vector-friendly index defaults suited to embedding storage and similarity search. +A Vector Database project gives you {{serverless-full}} operations with defaults and project settings aimed at embedding storage, {{infer}}, and similarity or hybrid search. -### Vector index mode by default +### Vector-optimized defaults and hardware profile -Indices in a Vector Database project use the [`index.mode: vectordb_document`](elasticsearch://reference/elasticsearch/mapping-reference/dense-vector.md#dense-vector-vectordb-document-mode) vector index mode automatically. It applies storage, indexing, and merge defaults tuned for similarity search on dense vectors, so you get efficient embedding storage and approximate [kNN](/solutions/search/vector/knn.md) search without configuring each setting yourself. +Indices in a Vector Database project use the [`index.mode: vectordb_document`](elasticsearch://reference/elasticsearch/mapping-reference/dense-vector.md#dense-vector-vectordb-document-mode) vector index mode automatically. It applies storage, indexing, and merge defaults tuned for similarity search on dense vectors, so you get efficient embedding storage and approximate [kNN](/solutions/search/vector/knn.md) search without configuring each setting yourself. The project hardware profile is also tuned for vector workloads. :::{tip} On other [deployment types](/deploy-manage/deploy.md), you can set the vector index mode explicitly when you create an index. @@ -34,23 +37,25 @@ The Vector Database project type favors workloads where you ingest and embed dat +### Access to {{infer}} + +Vector Database projects are set up for embedding workflows: generate vectors in {{es}} with managed models (for example through `semantic_text`), or store vectors you create yourself and attach the same model at query time. In-product setup guides walk through both paths. + ### Multi-tenant partitioning with slices For multi-tenant or multi-customer vector apps, you can use slices to partition an index so each tenant’s vectors are indexed and searched in isolation. Slice mode is opt-in and is not enabled automatically on Vector Database projects. -### Usage-based {{serverless-short}} operations +### Pricing designed for vector workloads Similar to other {{serverless-full}} projects, Elastic manages the infrastructure, scaling, and upgrades. You create a project, get an endpoint, and start indexing and querying without sizing nodes for vector RAM yourself. -Billing uses storage, searchable capacity, indexing volume, and infrastructure hours rather than the VCU model used by {{es-serverless}}. See [{{es}} Vector Database billing dimensions](/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md). +Billing uses storage, search, and indexing, rather than the compute-based VCU model used by {{es-serverless}} projects. Refer to [{{es}} Vector Database billing dimensions](/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md) for details. ## When to use this project type -Choose {{es}} Vector Database when you are building retrieval for AI features: a chatbot, RAG pipeline, recommendation engine, or similar workloads where your application consumes results. This provides out of the box vector-tuned defaults. - -Choose the [{{es}} project type](/solutions/elasticsearch-solution-project.md) when you are building a search application people interact with directly (a search bar, catalog, or knowledge base), or when you need broader search application tooling in {{kib}} such as [Playground](/solutions/elasticsearch-solution-project/playground.md), [Query Rules UI](/solutions/elasticsearch-solution-project/query-rules-ui.md), or [Agent Builder](/explore-analyze/ai-features/elastic-agent-builder.md). +Both the {{es}} Vector Database and the {{es}} project types support [vector search](/solutions/search/vector.md). Choose Vector Database when embeddings and similarity search are central to the workload. Choose the [{{es}} project type](/solutions/elasticsearch-solution-project.md) for general-purpose data storage and search, including mixed lexical, time series, and analytics workloads, or {{kib}} search tooling such as [Playground](/solutions/elasticsearch-solution-project/playground.md), [Query Rules UI](/solutions/elasticsearch-solution-project/query-rules-ui.md), and [Agent Builder](/explore-analyze/ai-features/elastic-agent-builder.md). You might prefer the {{es}} project type if you are an existing {{es}} or OpenSearch user. | Use case | Fit | Why | | --- | --- | --- | @@ -59,8 +64,12 @@ Choose the [{{es}} project type](/solutions/elasticsearch-solution-project.md) w | [Multimodal search](/solutions/search/vector/vector-search-use-cases.md#multimodal-search) | Strong | Search across images, audio, video, or text with embeddings from a multimodal model | | [Duplicate detection, fraud, and anomaly detection](/solutions/search/vector/vector-search-use-cases.md#duplicate-detection-fraud-and-anomaly-detection) | Strong | Compare embeddings to find near-duplicates, suspicious matches, or unusual patterns at scale | | [Long-term memory for LLMs](/solutions/search/vector/vector-search-use-cases.md#long-term-memory-for-llms) | Strong | Store facts, chat turns, or summaries so an assistant can retrieve relevant past context | -| Full-text or keyword search without vectors | Prefer the {{es}} project | The {{es}} project is built for lexical search, filters, and analytics on document-centric data | -| Log, event, or other time series search | Prefer {{es}} project | Choose the {{es}} project for continuously ingested logs and events, where general-purpose defaults fit write-heavy, frequently updated data | +| Full-text or keyword search without vectors | Prefer the {{es}} project | General-purpose defaults suit lexical search, filters, and document-centric analytics | +| Log, event, or other time series search | Prefer the {{es}} project | General-purpose defaults suit write-heavy, frequently updated time series data | + +:::{note} +Vector Database projects use vector index mode only. Time series (tsdb) and LogsDB index modes are not supported; use the {{es}} project type for those workloads. Data streams are supported when their backing indices use vector index mode. +::: ## Get started diff --git a/solutions/vector-database/get-started.md b/solutions/vector-database/get-started.md index e9f46e1b23..b81db04e9a 100644 --- a/solutions/vector-database/get-started.md +++ b/solutions/vector-database/get-started.md @@ -45,7 +45,7 @@ After you've created your project, note the {{es}} endpoint and API key from the When you create a new Vector Database project, the **Set up your Elasticsearch Vector Database** page includes two guided paths you can follow. Each path walks you through ingest and then search examples, with sample scripts you can run in a client of your choosing or run the examples directly in the Console. -Choose one: +The following setup guides are available in {{kib}}: | Guide | When to use | What you do | | --- | --- | --- | @@ -91,21 +91,16 @@ To walk through kNN search, follow [kNN search in {{es}}](/solutions/search/vect For more query options, see [Querying for search](/solutions/search/querying-for-search.md) and the [{{es-serverless}} API documentation]({{es-serverless-apis}}). ::::: - -:::::{step} Tune for your workload - -After you have a baseline that returns relevant results: - -* Review [vector search](/solutions/search/vector.md) concepts (field types, quantization, similarity) -* Adjust relevance with [ranking and reranking](/solutions/search/ranking.md) -* For RAG applications, see [RAG](/solutions/search/rag.md) -* Review project [Search Power](/deploy-manage/deploy/elastic-cloud/project-settings.md#elasticsearch-manage-project-search-power-settings) and [billing dimensions](/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md) for latency versus cost -::::: :::::: ## Next steps -* [Vector search in {{es}}](/solutions/search/vector.md): Concepts, field types, and query options +After you've learned how to ingest embeddings and return relevant results, dig deeper into how vector search works, improve ranking for your use case, and tune project settings for latency and cost. Review the following: + +* [Vector search in {{es}}](/solutions/search/vector.md): Concepts, field types, quantization, and query options +* [Ranking and reranking](/solutions/search/ranking.md): Improve relevance after you have a baseline +* [RAG](/solutions/search/rag.md): Patterns for grounding LLMs on retrieved context * [Vector search use cases](/solutions/search/vector/vector-search-use-cases.md): RAG, recommendations, multimodal search, and more -* [{{es}} Vector Database overview](/solutions/vector-database.md): When to choose this project type -* [{{es}} solution](/solutions/elasticsearch-solution-project.md): UI tools for broader search application development +* [Search Power](/deploy-manage/deploy/elastic-cloud/project-settings.md#elasticsearch-manage-project-search-power-settings) and [billing dimensions](/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md): Balance latency and cost for your project +* [{{es}} Vector Database overview](/solutions/vector-database.md): When to choose this project type versus the general-purpose {{es}} project +* [{{es}} solution](/solutions/elasticsearch-solution-project.md): General-purpose project type and search application UI tools From 28284d1d00ef06c51a8ed2623d4124f1e78147fb Mon Sep 17 00:00:00 2001 From: Vlada Chirmicci Date: Mon, 27 Jul 2026 17:14:17 +0100 Subject: [PATCH 03/13] Add post-review suggestions --- .../vector-database-billing-dimensions.md | 28 +++++++++---------- solutions/vector-database.md | 18 ++++-------- solutions/vector-database/get-started.md | 26 ++++++++--------- 3 files changed, 30 insertions(+), 42 deletions(-) diff --git a/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md b/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md index 526481e6fe..e4195a47f7 100644 --- a/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md +++ b/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md @@ -1,25 +1,25 @@ --- navigation_title: Vector Database projects applies_to: - serverless: preview + serverless: ga products: - id: cloud-serverless - id: elasticsearch description: >- - Learn how costs for Elasticsearch Vector Database Serverless projects are - calculated across storage, search, and indexing. + Learn how costs for Elasticsearch Vector Database Serverless projects + are calculated across storage, search, indexing, and infrastructure. --- # {{es}} Vector Database billing dimensions [vector-database-billing] -{{es}} Vector Database projects on {{serverless-full}} are priced based on consumption across storage, search, and indexing. +{{es}} Vector Database projects on {{serverless-full}} are priced based on consumption across storage, search, indexing, and infrastructure. Your monthly bill is calculated based on these components: * **Storage** — Measured by the total amount of data stored in your project, in GB. -* **Search** — Measured by the resources allocated to keep your data searchable. Allocation is controlled by your project's [Search Power](/deploy-manage/deploy/elastic-cloud/project-settings.md#elasticsearch-manage-project-search-power-settings) setting. +* **Search** — Measured by the resources allocated to keep your data hot. Allocation is controlled by your project's [Search Power](/deploy-manage/deploy/elastic-cloud/project-settings.md#elasticsearch-manage-project-search-power-settings) setting. * **Indexing** — Measured by the volume of data you index into your project over the course of a month, in GB. - +* **Infrastructure fee** — Measured by the number of hours your project is running. For current rates, refer to the [Cloud Pricing Table](https://cloud.elastic.co/cloud-pricing-table?productType=serverless) or the [{{serverless-full}} pricing](https://www.elastic.co/pricing/serverless) pages. @@ -31,28 +31,26 @@ You are charged per GB of data stored in the project. Storage charges apply for ## Search [vector-database-billing-search] -Search charges cover the resources used to keep your data searchable. All stored data is boosted by default, which means it is searchable and contributes to Search charges. +Search charges cover the resources used to keep your data hot for active searches. All stored data is hot by default, which means it actively contributes to Search charges. [Search Power](/deploy-manage/deploy/elastic-cloud/project-settings.md#elasticsearch-manage-project-search-power-settings) controls how many baseline resources are kept ready for your data and the maximum resources the system can allocate under search load. Increasing Search Power improves query performance and increases search charges. Decreasing Search Power reduces provisioned resources and cost, with more variable query latency under load. ## Indexing [vector-database-billing-indexing] -Indexing charges are based on the volume of data written to your project, measured in GB for the billing period. This uses the same definition of billable indexing bytes as other {{serverless-short}} offerings. +Indexing charges are based on the volume of data written to your project, measured in GB for the billing period. This uses the same definition of billable indexing bytes as other Serverless offerings. + +## Infrastructure [vector-database-billing-infrastructure] + +You are billed an hourly infrastructure fee for every hour that your project is running. ## Managing Vector Database costs [vector-database-billing-managing-costs] -Vector Database costs follow your storage footprint, search resource allocation, and indexing volume. To balance performance with spend, adjust the controls described in this section. +Vector Database costs follow your storage footprint, search resource allocation, indexing volume, and the hours your project is available. To balance performance with spend, adjust the controls described in this section. ### Search Power setting [vector-database-billing-search-power-setting] [Search Power](/deploy-manage/deploy/elastic-cloud/project-settings.md#elasticsearch-manage-project-search-power-settings) controls search performance and the resources allocated to boosted data. Increase Search Power when you need more consistent latency and throughput. Decrease it when you want to reduce Search charges and can accept more variable latency. -### Inactivity API [vector-database-billing-inactivity-api] - -Boosted versus unboosted is not a customer-facing index setting. All storage is boosted by default. To reduce Search charges for data you do not need to keep searchable, use the Inactivity API. - - - ## Related billing dimensions [vector-database-billing-related] Shared {{serverless-short}} add-ons such as [data out](serverless-project-billing-dimensions.md#general-serverless-billing-data-out) and [support](serverless-project-billing-dimensions.md#general-serverless-billing-support) may also appear on your bill. See [](serverless-project-billing-dimensions.md). diff --git a/solutions/vector-database.md b/solutions/vector-database.md index cfca7dbbe5..3e9b87c3d4 100644 --- a/solutions/vector-database.md +++ b/solutions/vector-database.md @@ -1,7 +1,7 @@ --- navigation_title: Vector Database project applies_to: - serverless: preview + serverless: ga description: >- The Elasticsearch Vector Database project type on Elastic Cloud Serverless is optimized for vector workloads, with vector-tuned defaults, hardware profile, @@ -17,7 +17,7 @@ The {{es}} Vector Database {{serverless-short}} project type is optimized for ve [Vector search](/solutions/search/vector.md) uses the same query APIs in both project types. -Use it when embeddings and similarity search are central to your application, for example RAG, recommendations, semantic retrieval, or multimodal search. You can run [semantic](/solutions/search/semantic-search.md) and [hybrid](/solutions/search/hybrid-search.md) search in the same project. +Use it when embeddings and similarity search are central to your application, for example RAG, recommendations, [semantic search](/solutions/search/semantic-search.md), [hybrid search](/solutions/search/hybrid-search.md), or multimodal search. ## What you get @@ -41,17 +41,11 @@ The Vector Database project type favors workloads where you ingest and embed dat Vector Database projects are set up for embedding workflows: generate vectors in {{es}} with managed models (for example through `semantic_text`), or store vectors you create yourself and attach the same model at query time. In-product setup guides walk through both paths. -### Multi-tenant partitioning with slices - -For multi-tenant or multi-customer vector apps, you can use slices to partition an index so each tenant’s vectors are indexed and searched in isolation. Slice mode is opt-in and is not enabled automatically on Vector Database projects. - - - ### Pricing designed for vector workloads Similar to other {{serverless-full}} projects, Elastic manages the infrastructure, scaling, and upgrades. You create a project, get an endpoint, and start indexing and querying without sizing nodes for vector RAM yourself. -Billing uses storage, search, and indexing, rather than the compute-based VCU model used by {{es-serverless}} projects. Refer to [{{es}} Vector Database billing dimensions](/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md) for details. +Billing uses storage, search, indexing, and infrastructure, rather than the compute-based VCU model used by {{es-serverless}} projects. Refer to [{{es}} Vector Database billing dimensions](/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md) for details. ## When to use this project type @@ -59,11 +53,11 @@ Both the {{es}} Vector Database and the {{es}} project types support [vector sea | Use case | Fit | Why | | --- | --- | --- | -| [RAG and question answering](/solutions/search/vector/vector-search-use-cases.md#rag-and-question-answering-on-your-own-data) | Strong | Retrieve passages from documents, wikis, tickets, or knowledge bases and pass them to an LLM for assistants, support bots, and cited answers | -| [Discovery and recommendations](/solutions/search/vector/vector-search-use-cases.md#discovery-and-recommendations) | Strong | Find related products, articles, or other items by similarity when keywords alone are not enough | +| [RAG and question answering](/solutions/search/vector/vector-search-use-cases.md#rag-and-question-answering-on-your-own-data) | Strong | Retrieve passages from documents, wikis, tickets, or knowledge bases and pass them to an LLM. Hybrid search combines semantic similarity with keyword matching when queries mix natural language with exact terms, IDs, or product names | +| [Discovery and recommendations](/solutions/search/vector/vector-search-use-cases.md#discovery-and-recommendations) | Strong | Find related products, articles, or other items by similarity when keywords alone are not enough. Use hybrid ranking when you also need lexical or attribute matches in the same result set | | [Multimodal search](/solutions/search/vector/vector-search-use-cases.md#multimodal-search) | Strong | Search across images, audio, video, or text with embeddings from a multimodal model | | [Duplicate detection, fraud, and anomaly detection](/solutions/search/vector/vector-search-use-cases.md#duplicate-detection-fraud-and-anomaly-detection) | Strong | Compare embeddings to find near-duplicates, suspicious matches, or unusual patterns at scale | -| [Long-term memory for LLMs](/solutions/search/vector/vector-search-use-cases.md#long-term-memory-for-llms) | Strong | Store facts, chat turns, or summaries so an assistant can retrieve relevant past context | +| [Long-term memory for LLMs](/solutions/search/vector/vector-search-use-cases.md#long-term-memory-for-llms) | Strong | Store facts, chat turns, or summaries so an assistant can retrieve relevant past context by meaning, optionally combined with keyword filters on metadata | | Full-text or keyword search without vectors | Prefer the {{es}} project | General-purpose defaults suit lexical search, filters, and document-centric analytics | | Log, event, or other time series search | Prefer the {{es}} project | General-purpose defaults suit write-heavy, frequently updated time series data | diff --git a/solutions/vector-database/get-started.md b/solutions/vector-database/get-started.md index b81db04e9a..62dabbb42b 100644 --- a/solutions/vector-database/get-started.md +++ b/solutions/vector-database/get-started.md @@ -4,7 +4,7 @@ description: >- Create an Elasticsearch Vector Database project on Elastic Cloud Serverless, ingest embeddings, and run your first vector or semantic searches. applies_to: - serverless: preview + serverless: ga products: - id: elasticsearch - id: cloud-serverless @@ -59,8 +59,13 @@ You can also skip the setup guide and continue with the steps below. Use the approach that matches how you create embeddings. ::::{dropdown} Generate embeddings from your content -Create an index with a [`semantic_text`](elasticsearch://reference/elasticsearch/mapping-reference/semantic-text.md) field, then ingest documents so {{es}} generates embeddings automatically. The default model is [`jina-embeddings-v3`](/explore-analyze/machine-learning/nlp/ml-nlp-jina.md#jina-embeddings-v3), integrated through the {{es}} {{infer}} pipeline. +You can generate embeddings as part of the ingestion workflow instead of creating them in advance. +For text content, the recommended approach is to map the target field as [`semantic_text`](elasticsearch://reference/elasticsearch/mapping-reference/semantic-text.md). When you ingest documents, {{es}} uses the configured {{infer}} endpoint to generate and store embeddings automatically. You can check which models are used by default and learn how to change them in [Configure {{infer}} endpoints documentation](elasticsearch://reference/elasticsearch/mapping-reference/semantic-text-setup-configuration.md#configure-inference-endpoints). + +For more control over the {{infer}} and ingestion workflow, you can also generate embeddings with an {{infer}} processor in an ingest pipeline and store them in a vector field. For more information, see [](/solutions/search/semantic-search/semantic-search-elser-ingest-pipelines.md). + +To walk through mapping a `semantic_text` field, ingesting sample content, and running a hybrid query, follow the [semantic search quickstart](/solutions/search/get-started/semantic-search.md). :::: ::::{dropdown} Store your existing embeddings @@ -73,21 +78,12 @@ Use a language [client](/reference/elasticsearch-clients/index.md) or the [Bulk] ::::: :::::{step} Search your data -Use the approach that matches how you ingested embeddings. +Match your query to the vector field type in your mapping. For details on which queries each field type supports, see [Field types and queries](/solutions/search/vector.md#vector-queries-and-field-types). -::::{dropdown} Search embeddings generated from your content -Query with natural language. {{es}} embeds the query and returns the most relevant matches. You can also run a hybrid query that combines semantic and lexical ranking. - -To walk through hybrid search with `semantic_text`, follow [Hybrid search with `semantic_text`](/solutions/search/hybrid-semantic-text.md). For query and retrieval options, see [Search and retrieve `semantic_text` fields](elasticsearch://reference/elasticsearch/mapping-reference/semantic-text-search-retrieval.md). -:::: +To walk through common patterns, refer to the following pages: -::::{dropdown} Search your existing embeddings -Run a kNN search against your `dense_vector` field, or combine kNN with lexical search for hybrid retrieval. - -Queries need to be converted into vectors the same way your documents were. Specify the model you used to generate your embeddings so {{es}} can apply it to your queries too. - -To walk through kNN search, follow [kNN search in {{es}}](/solutions/search/vector/knn.md). For combining vector and lexical ranking, see [Hybrid search](/solutions/search/hybrid-search.md). -:::: +* [kNN search in {{es}}](/solutions/search/vector/knn.md) +* [Hybrid search](/solutions/search/hybrid-search.md) For more query options, see [Querying for search](/solutions/search/querying-for-search.md) and the [{{es-serverless}} API documentation]({{es-serverless-apis}}). ::::: From 0807c3864226fff7a7d3d8a67eddbb1da1c5800d Mon Sep 17 00:00:00 2001 From: Vlada Chirmicci Date: Wed, 29 Jul 2026 15:35:24 +0100 Subject: [PATCH 04/13] Another round of reviews + search power details --- .../billing/vector-database-billing-dimensions.md | 14 ++++++++++---- .../deploy/elastic-cloud/project-settings.md | 1 + 2 files changed, 11 insertions(+), 4 deletions(-) diff --git a/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md b/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md index e4195a47f7..e120e407dd 100644 --- a/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md +++ b/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md @@ -17,7 +17,7 @@ description: >- Your monthly bill is calculated based on these components: * **Storage** — Measured by the total amount of data stored in your project, in GB. -* **Search** — Measured by the resources allocated to keep your data hot. Allocation is controlled by your project's [Search Power](/deploy-manage/deploy/elastic-cloud/project-settings.md#elasticsearch-manage-project-search-power-settings) setting. +* **Search** — Measured by reserved search capacity for your stored data. Capacity is controlled by your project's [Search Power](/deploy-manage/deploy/elastic-cloud/project-settings.md#elasticsearch-manage-project-search-power-settings) setting and scales with how much data you store. * **Indexing** — Measured by the volume of data you index into your project over the course of a month, in GB. * **Infrastructure fee** — Measured by the number of hours your project is running. @@ -31,9 +31,11 @@ You are charged per GB of data stored in the project. Storage charges apply for ## Search [vector-database-billing-search] -Search charges cover the resources used to keep your data hot for active searches. All stored data is hot by default, which means it actively contributes to Search charges. +Search charges cover the search capacity reserved for your stored data. All stored data is searchable by default, which means it actively contributes to Search charges. -[Search Power](/deploy-manage/deploy/elastic-cloud/project-settings.md#elasticsearch-manage-project-search-power-settings) controls how many baseline resources are kept ready for your data and the maximum resources the system can allocate under search load. Increasing Search Power improves query performance and increases search charges. Decreasing Search Power reduces provisioned resources and cost, with more variable query latency under load. +[Search Power](/deploy-manage/deploy/elastic-cloud/project-settings.md#elasticsearch-manage-project-search-power-settings) determines how much capacity is reserved. At Search Power **100**, enough capacity is reserved so that 100% of your project data remains available for low-latency search. Higher Search Power values reserve more capacity in proportion to that baseline: for example, **200** reserves about twice as much as **100**. + +Increasing Search Power reserves more capacity and raises search charges. Running more queries does not increase search charges. For current rates, refer to the [Cloud Pricing Table](https://cloud.elastic.co/cloud-pricing-table?productType=serverless). ## Indexing [vector-database-billing-indexing] @@ -49,8 +51,12 @@ Vector Database costs follow your storage footprint, search resource allocation, ### Search Power setting [vector-database-billing-search-power-setting] -[Search Power](/deploy-manage/deploy/elastic-cloud/project-settings.md#elasticsearch-manage-project-search-power-settings) controls search performance and the resources allocated to boosted data. Increase Search Power when you need more consistent latency and throughput. Decrease it when you want to reduce Search charges and can accept more variable latency. +[Search Power](/deploy-manage/deploy/elastic-cloud/project-settings.md#elasticsearch-manage-project-search-power-settings) reserves search capacity for your project. Start at the default of **100**, measure latency and throughput for your workload, then increase Search Power if you need more capacity. Higher Search Power increases search charges in proportion to the setting and your stored data volume. + +For high availability, you can increase Search Power to **200**. That setting reserves enough capacity for two full copies of your project data to remain available for low-latency search. ## Related billing dimensions [vector-database-billing-related] Shared {{serverless-short}} add-ons such as [data out](serverless-project-billing-dimensions.md#general-serverless-billing-data-out) and [support](serverless-project-billing-dimensions.md#general-serverless-billing-support) may also appear on your bill. See [](serverless-project-billing-dimensions.md). + +If you use {{cps}}, additional charges may apply when the feature becomes generally available. For how those charges are calculated, refer to [ {{cps}} Billing](/deploy-manage/cross-project-search-config.md#cps-billing). diff --git a/deploy-manage/deploy/elastic-cloud/project-settings.md b/deploy-manage/deploy/elastic-cloud/project-settings.md index 49017c2755..3543bc70d4 100644 --- a/deploy-manage/deploy/elastic-cloud/project-settings.md +++ b/deploy-manage/deploy/elastic-cloud/project-settings.md @@ -54,6 +54,7 @@ $$$elasticsearch-manage-project-search-power-settings$$$ | Setting | Description | Project Type | | :--- | :--- | :--- | | **Search Power** | Search Power controls the amount of baseline resources kept ready for your data and the maximum amount of resources the system can allocate to your project under high search load.

Increase **Search Power** to add resources for improved query throughput and performance as load increases. Decrease it to reduce the amount of provisioned resources, which lowers both active search costs and the reduced rate billed during idle periods, at the cost of more variable query latency. The On-demand setting provides the lowest baseline and idle cost.

Choose from three Search Power settings:

**On-demand:** Low cost option. Query latency is variable and maximum throughput is limited, particularly during demand spikes.

**Performant:** Delivers low _average_ query latency with no practical query throughput limits to meet varying demand.

**High availability:** Maintains consistent low query latency across most search workloads, from steady baseline to peak demand.
| {{es}} | +| **Search Power** | Search Power reserves search capacity for your project. Unlike {{es}} projects, capacity is not scaled automatically based on query load.

The default value is **100**, which reserves enough search resources so that 100% of your project data remains available for low-latency search. You can increase Search Power in increments of 100. Higher values reserve proportionally more capacity: for example, **200** reserves about twice the capacity of **100**.
| {{es}} Vector Database | | **Search Boost Window** | Non-time series data is always considered search-ready. The **Search Boost Window** determines the volume of time series project data that will be considered search-ready.

Increasing the window results in a bigger portion of time series project data included in the total search-ready data volume.
| {{es}} | | **Data Retention** | Data retention policies determine how long your project data is retained.
In {{serverless-full}} data retention policies are configured through [data streams](../../../manage-data/lifecycle/data-stream.md) and you can [specify different retention periods](../../../manage-data/lifecycle/data-stream/tutorial-update-existing-data-stream.md#set-lifecycle) for specific data streams in your project.

{{elastic-sec}} has two additional configuration settings that can be configured to manage your data retention.

**Maximum data retention period**

When enabled, this setting determines the maximum length of time that data can be retained in any data streams of this project.

Editing this setting replaces the data retention set for all data streams of the project that have a longer data retention defined. Data older than the new maximum retention period that you set is permanently deleted.

**Default data retention period**

When enabled, this setting determines the default retention period that is automatically applied to all data streams in your project that do not have a custom retention period already set.
|Elasticsearch
Observability
Security | | **Project features** | Controls [feature tiers and add-on options](../../../deploy-manage/deploy/elastic-cloud/project-settings.md#project-features-add-ons) for your {{elastic-sec}} project. | Security | From fc037f4b3d2ea524a5ce7e0ec0e5601eeff06980 Mon Sep 17 00:00:00 2001 From: Vlada Chirmicci Date: Wed, 29 Jul 2026 15:46:15 +0100 Subject: [PATCH 05/13] Substitute CPS architecture image with a mermaid diagram The diagram also includes a new node to represent Vector Database project types. --- deploy-manage/cross-project-search-config.md | 11 ++++++++++- 1 file changed, 10 insertions(+), 1 deletion(-) diff --git a/deploy-manage/cross-project-search-config.md b/deploy-manage/cross-project-search-config.md index 6399ea4ce0..0179f154f7 100644 --- a/deploy-manage/cross-project-search-config.md +++ b/deploy-manage/cross-project-search-config.md @@ -75,7 +75,16 @@ For most deployments, we recommend creating a dedicated **overview project** tha In this architecture, you create a new, empty project and link existing projects to it. You run all cross-project searches from the new overview project, while your actual active projects continue to operate independently. The linked ("spoke") projects are not linked to each other. -![Overview project architecture for cross-project search](images/serverless-cross-project-search-arch.svg) +```mermaid +flowchart TB + O["Overview project
Origin (empty hub)"] + O --> S["Security project
Linked (data)"] + O --> Obs["Observability project
Linked (data)"] + O --> E["Elasticsearch project
Linked (data)"] + O --> V["Vector Database project
Linked (data)"] +``` + +Searches run from the overview project across all linked projects. Linked projects operate independently and are not linked to each other. You can link any combination of compatible projects. The overview project becomes a central point for broad searches, dashboards, and investigations, without affecting your existing setup. From cb31b4b2cae2806bd9604a15eeb156b7b1b3d364 Mon Sep 17 00:00:00 2001 From: Vlada Chirmicci Date: Thu, 30 Jul 2026 15:17:26 +0100 Subject: [PATCH 06/13] Adding a few more edits + updatign the user roles table --- .../billing/vector-database-billing-dimensions.md | 12 ++++++------ deploy-manage/images/serverless-vectordb-badge.svg | 14 ++++++++++++++ .../users-roles/cloud-organization/user-roles.md | 6 +++--- solutions/vector-database.md | 2 +- solutions/vector-database/get-started.md | 2 +- 5 files changed, 25 insertions(+), 11 deletions(-) create mode 100644 deploy-manage/images/serverless-vectordb-badge.svg diff --git a/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md b/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md index e120e407dd..5f26a289b8 100644 --- a/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md +++ b/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md @@ -7,18 +7,18 @@ products: - id: elasticsearch description: >- Learn how costs for Elasticsearch Vector Database Serverless projects - are calculated across storage, search, indexing, and infrastructure. + are calculated across storage, search, ingest, and infrastructure. --- # {{es}} Vector Database billing dimensions [vector-database-billing] -{{es}} Vector Database projects on {{serverless-full}} are priced based on consumption across storage, search, indexing, and infrastructure. +{{es}} Vector Database projects on {{serverless-full}} are priced based on consumption across storage, search, ingest, and infrastructure. Your monthly bill is calculated based on these components: * **Storage** — Measured by the total amount of data stored in your project, in GB. * **Search** — Measured by reserved search capacity for your stored data. Capacity is controlled by your project's [Search Power](/deploy-manage/deploy/elastic-cloud/project-settings.md#elasticsearch-manage-project-search-power-settings) setting and scales with how much data you store. -* **Indexing** — Measured by the volume of data you index into your project over the course of a month, in GB. +* **Ingest** — Measured by the volume of data you ingest into your project over the course of a month, in GB. * **Infrastructure fee** — Measured by the number of hours your project is running. For current rates, refer to the [Cloud Pricing Table](https://cloud.elastic.co/cloud-pricing-table?productType=serverless) or the [{{serverless-full}} pricing](https://www.elastic.co/pricing/serverless) pages. @@ -37,9 +37,9 @@ Search charges cover the search capacity reserved for your stored data. All stor Increasing Search Power reserves more capacity and raises search charges. Running more queries does not increase search charges. For current rates, refer to the [Cloud Pricing Table](https://cloud.elastic.co/cloud-pricing-table?productType=serverless). -## Indexing [vector-database-billing-indexing] +## Ingest [vector-database-billing-ingest] -Indexing charges are based on the volume of data written to your project, measured in GB for the billing period. This uses the same definition of billable indexing bytes as other Serverless offerings. +Ingest charges are based on the volume of data written to your project, measured in GB for the billing period. This uses the same definition of billable ingested bytes as other Serverless offerings. ## Infrastructure [vector-database-billing-infrastructure] @@ -47,7 +47,7 @@ You are billed an hourly infrastructure fee for every hour that your project is ## Managing Vector Database costs [vector-database-billing-managing-costs] -Vector Database costs follow your storage footprint, search resource allocation, indexing volume, and the hours your project is available. To balance performance with spend, adjust the controls described in this section. +Vector Database costs follow your storage footprint, search resource allocation, ingest volume, and the hours your project is available. To balance performance with spend, adjust the controls described in this section. ### Search Power setting [vector-database-billing-search-power-setting] diff --git a/deploy-manage/images/serverless-vectordb-badge.svg b/deploy-manage/images/serverless-vectordb-badge.svg new file mode 100644 index 0000000000..504844c41b --- /dev/null +++ b/deploy-manage/images/serverless-vectordb-badge.svg @@ -0,0 +1,14 @@ + + + + + + + + + + + + + + diff --git a/deploy-manage/users-roles/cloud-organization/user-roles.md b/deploy-manage/users-roles/cloud-organization/user-roles.md index 442e3ba8e9..c2b8b10dcf 100644 --- a/deploy-manage/users-roles/cloud-organization/user-roles.md +++ b/deploy-manage/users-roles/cloud-organization/user-roles.md @@ -126,9 +126,9 @@ You can optionally [create custom roles in a project](/deploy-manage/users-roles | Role name | {{ecloud}} access | Project access | Availability | | --- | --- | --- | --- | -| Admin | Has full access to project management, properties, and security privileges. | Superuser role privileges | [![{{es}}](/deploy-manage/images/serverless-es-badge.svg "")](../../../solutions/search.md)[![{{observability}}](/deploy-manage/images/serverless-obs-badge.svg "")](../../../solutions/observability.md)[![Security](/deploy-manage/images/serverless-sec-badge.svg "")](../../../solutions/security.md) | -| Developer | Manage project settings. | Creates API keys, indices, data streams, adds connectors, and builds visualizations. | [![{{es}}](/deploy-manage/images/serverless-es-badge.svg "")](../../../solutions/search.md) | -| Viewer | Has read-only access to project details. | Has read-only access to project data and features. | [![{{es}}](/deploy-manage/images/serverless-es-badge.svg "")](../../../solutions/search.md)[![{{observability}}](/deploy-manage/images/serverless-obs-badge.svg "")](../../../solutions/observability.md)[![Security](/deploy-manage/images/serverless-sec-badge.svg "")](../../../solutions/security.md) | +| Admin | Has full access to project management, properties, and security privileges. | Superuser role privileges | [![{{es}}](/deploy-manage/images/serverless-es-badge.svg "")](../../../solutions/search.md)[![Vector Database](/deploy-manage/images/serverless-vectordb-badge.svg "")](../../../solutions/vector-database.md)[![{{observability}}](/deploy-manage/images/serverless-obs-badge.svg "")](../../../solutions/observability.md)[![Security](/deploy-manage/images/serverless-sec-badge.svg "")](../../../solutions/security.md) | +| Developer | Manage project settings. | Creates API keys, indices, data streams, adds connectors, and builds visualizations. | [![{{es}}](/deploy-manage/images/serverless-es-badge.svg "")](../../../solutions/search.md)[![Vector Database](/deploy-manage/images/serverless-vectordb-badge.svg "")](../../../solutions/vector-database.md) | +| Viewer | Has read-only access to project details. | Has read-only access to project data and features. | [![{{es}}](/deploy-manage/images/serverless-es-badge.svg "")](../../../solutions/search.md)[![Vector Database](/deploy-manage/images/serverless-vectordb-badge.svg "")](../../../solutions/vector-database.md)[![{{observability}}](/deploy-manage/images/serverless-obs-badge.svg "")](../../../solutions/observability.md)[![Security](/deploy-manage/images/serverless-sec-badge.svg "")](../../../solutions/security.md) | | Editor | Configures all {{observability}} or Security projects. | Has read-only access to data indices. Has full access to all project features. | [![{{observability}}](/deploy-manage/images/serverless-obs-badge.svg "")](../../../solutions/observability.md)[![Security](/deploy-manage/images/serverless-sec-badge.svg "")](../../../solutions/security.md) | | Tier 1 analyst | Viewer | Ideal for initial alert triage. General read access, can create dashboards and visualizations. | [![Security](/deploy-manage/images/serverless-sec-badge.svg "")](../../../solutions/security.md) | | Tier 2 analyst | Viewer | Ideal for alert triage and beginning the investigation process. Can create cases. | [![Security](/deploy-manage/images/serverless-sec-badge.svg "")](../../../solutions/security.md) | diff --git a/solutions/vector-database.md b/solutions/vector-database.md index 3e9b87c3d4..4f8f753cd4 100644 --- a/solutions/vector-database.md +++ b/solutions/vector-database.md @@ -45,7 +45,7 @@ Vector Database projects are set up for embedding workflows: generate vectors in Similar to other {{serverless-full}} projects, Elastic manages the infrastructure, scaling, and upgrades. You create a project, get an endpoint, and start indexing and querying without sizing nodes for vector RAM yourself. -Billing uses storage, search, indexing, and infrastructure, rather than the compute-based VCU model used by {{es-serverless}} projects. Refer to [{{es}} Vector Database billing dimensions](/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md) for details. +Billing uses storage, search, ingest, and infrastructure, rather than the compute-based VCU model used by {{es-serverless}} projects. Refer to [{{es}} Vector Database billing dimensions](/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md) for details. ## When to use this project type diff --git a/solutions/vector-database/get-started.md b/solutions/vector-database/get-started.md index 62dabbb42b..19e8ab1d5e 100644 --- a/solutions/vector-database/get-started.md +++ b/solutions/vector-database/get-started.md @@ -25,7 +25,7 @@ If you're looking for an introduction to the {{stack}} or the {{es}} product, go There are two options to create serverless projects: -* If you're a new user, [sign up for a free 14-day trial](https://cloud.elastic.co/serverless-registration). For more information about {{ecloud}} trials, check out [Trial information](/deploy-manage/deploy/elastic-cloud/create-an-organization.md#general-sign-up-trial-what-is-included-in-my-trial). +* If you're a new user, [sign up for a free 14-day trial](https://cloud.elastic.co/serverless-registration?onboarding_token=vector). For more information about {{ecloud}} trials, check out [Trial information](/deploy-manage/deploy/elastic-cloud/create-an-organization.md#general-sign-up-trial-what-is-included-in-my-trial). * If you're an existing customer, [log in to {{ecloud}}](https://cloud.elastic.co/login) and do the following: 1. Select **Create project** from the **Serverless projects** panel. 2. Select **Next** from the **Vector Database** panel. From f743c2139c7e7126a7050dfb6858f01dc514a7e2 Mon Sep 17 00:00:00 2001 From: Vlada Chirmicci Date: Fri, 31 Jul 2026 17:23:57 +0100 Subject: [PATCH 07/13] Updating more entry points for the VectorDB docs --- get-started/introduction.md | 3 ++- get-started/versioning-availability.md | 2 +- solutions/index.md | 2 +- 3 files changed, 4 insertions(+), 3 deletions(-) diff --git a/get-started/introduction.md b/get-started/introduction.md index 9cd36eaf5a..c732081476 100644 --- a/get-started/introduction.md +++ b/get-started/introduction.md @@ -17,7 +17,8 @@ Whether you're building a search experience, monitoring your infrastructure, or | Your need | Recommended solution | Best for | |-----------|-------------------|----------| -| Build powerful, scalable searches to quickly search, analyze, and visualize large amounts of data for real-time insights| [{{es}}](/solutions/elasticsearch-solution-project.md)
• [Get started](/solutions/elasticsearch-solution-project/get-started.md)| Developers, architects, data engineers | +| Build powerful, scalable search and analytics applications across structured data, logs, metrics, documents, and vectors as part of a broader {{stack}} | [{{es}}](/solutions/elasticsearch-solution-project.md)
• [Get started](/solutions/elasticsearch-solution-project/get-started.md)| Developers, architects, data engineers | +| Build embedding-driven workloads such as semantic search, RAG, and AI-powered retrieval | [{{es}} Vector Database](/solutions/vector-database.md) {applies_to}`serverless:`
• [Get started](/solutions/vector-database/get-started.md) {applies_to}`serverless:` | Developers building AI retrieval and similarity search applications | | Observe and monitor system health and performance, or send telemetry data | [Elastic {{observability}}](/solutions/observability.md)
• [Get started](/solutions/observability/get-started.md) | DevOps, SREs, IT operations | | Monitor data for anomalous activity, detect, prevent, and respond to security incidents | [{{elastic-sec}}](/solutions/security.md)
• [Get started](/solutions/security/get-started.md)| SOC teams, security analysts, IT security admins | diff --git a/get-started/versioning-availability.md b/get-started/versioning-availability.md index edf36797b3..15608dcd5a 100644 --- a/get-started/versioning-availability.md +++ b/get-started/versioning-availability.md @@ -139,7 +139,7 @@ Available features can differ based on deployment type, product lifecycle stage, | {{stack}} versions | [{{stack}}](/get-started/the-stack.md) version 9.0.0 and later, including 9.1.0 | | Deployment types | [{{serverless-full}}](/deploy-manage/deploy/elastic-cloud/serverless.md), [{{ech}}](/deploy-manage/deploy/elastic-cloud/cloud-hosted.md), [{{ece}}](/deploy-manage/deploy/cloud-enterprise.md), [{{eck}}](/deploy-manage/deploy/cloud-on-k8s.md), and [Self-managed deployments](/deploy-manage/deploy/self-managed.md) | | Orchestrator versions | [{{ece}}](/deploy-manage/deploy/cloud-enterprise.md) 4.0.0 and later, [{{eck}}](/deploy-manage/deploy/cloud-on-k8s.md) 3.0.0 and later | -| Serverless project types | {{es}}, Elastic {{observability}}, and {{elastic-sec}} | +| Serverless project types | {{es}}, {{es}} Vector Database, Elastic {{observability}}, and {{elastic-sec}} | ## Find docs for your product version diff --git a/solutions/index.md b/solutions/index.md index fd08e6ea9e..75df33dd62 100644 --- a/solutions/index.md +++ b/solutions/index.md @@ -28,7 +28,7 @@ Elastic helps you build applications for three main use cases: search, observabi | Protecting against threats | [Security solution](/solutions/security.md) | Detect and respond to security threats | ::::{tip} -Not sure which to choose? Start with the {{es}} solution, if you don't need the additional features in {{product.observability}} or {{product.security}}. +Not sure which to choose? Start with the {{es}} solution for general-purpose search and analytics across structured data, logs, metrics, documents, and vectors. On {{serverless-full}}, choose [{{es}} Vector Database](/solutions/vector-database.md) when embeddings drive the workload. Choose {{product.observability}} or {{product.security}} when you need those solution features. :::: ## About solutions and project types From e466b4362d31a5dcca4bc8168389e0606398d9d7 Mon Sep 17 00:00:00 2001 From: Vlada Chirmicci Date: Tue, 4 Aug 2026 12:20:22 +0100 Subject: [PATCH 08/13] Add ML nodes limitation --- solutions/vector-database.md | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/solutions/vector-database.md b/solutions/vector-database.md index 4f8f753cd4..72635a8502 100644 --- a/solutions/vector-database.md +++ b/solutions/vector-database.md @@ -49,7 +49,11 @@ Billing uses storage, search, ingest, and infrastructure, rather than the comput ## When to use this project type -Both the {{es}} Vector Database and the {{es}} project types support [vector search](/solutions/search/vector.md). Choose Vector Database when embeddings and similarity search are central to the workload. Choose the [{{es}} project type](/solutions/elasticsearch-solution-project.md) for general-purpose data storage and search, including mixed lexical, time series, and analytics workloads, or {{kib}} search tooling such as [Playground](/solutions/elasticsearch-solution-project/playground.md), [Query Rules UI](/solutions/elasticsearch-solution-project/query-rules-ui.md), and [Agent Builder](/explore-analyze/ai-features/elastic-agent-builder.md). You might prefer the {{es}} project type if you are an existing {{es}} or OpenSearch user. +Both the {{es}} Vector Database and the {{es}} project types support [vector search](/solutions/search/vector.md). + +Choose Vector Database when embeddings and similarity search are central to the workload. For those projects, generate embeddings with managed models through the [Elastic Inference Service](/explore-analyze/elastic-inference/eis.md) or a third-party inference endpoint; you can't deploy custom ML models for in-cluster inference. + +Choose the [{{es}} project type](/solutions/elasticsearch-solution-project.md) when you need general-purpose data storage and search, including mixed lexical, time series, and analytics workloads, {{kib}} search tooling such as [Playground](/solutions/elasticsearch-solution-project/playground.md), [Query Rules UI](/solutions/elasticsearch-solution-project/query-rules-ui.md), and [Agent Builder](/explore-analyze/ai-features/elastic-agent-builder.md), or the ability to run custom models on ML nodes. You might also prefer the {{es}} project type if you are an existing {{es}} or OpenSearch user. | Use case | Fit | Why | | --- | --- | --- | From 05ff4018804432334005b3b6ea6c4a205cd2e2aa Mon Sep 17 00:00:00 2001 From: Vlada Chirmicci Date: Tue, 4 Aug 2026 15:29:42 +0100 Subject: [PATCH 09/13] Add SP limit value --- .../billing/vector-database-billing-dimensions.md | 2 +- deploy-manage/deploy/elastic-cloud/project-settings.md | 2 +- solutions/vector-database.md | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-) diff --git a/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md b/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md index 5f26a289b8..2d86d9a76a 100644 --- a/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md +++ b/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md @@ -51,7 +51,7 @@ Vector Database costs follow your storage footprint, search resource allocation, ### Search Power setting [vector-database-billing-search-power-setting] -[Search Power](/deploy-manage/deploy/elastic-cloud/project-settings.md#elasticsearch-manage-project-search-power-settings) reserves search capacity for your project. Start at the default of **100**, measure latency and throughput for your workload, then increase Search Power if you need more capacity. Higher Search Power increases search charges in proportion to the setting and your stored data volume. +[Search Power](/deploy-manage/deploy/elastic-cloud/project-settings.md#elasticsearch-manage-project-search-power-settings) reserves search capacity for your project. Start at the default of **100**, measure latency and throughput for your workload, then increase Search Power if you need more capacity. Higher Search Power increases search charges in proportion to the setting and your stored data volume. You can increase Search Power up to **1900**. For high availability, you can increase Search Power to **200**. That setting reserves enough capacity for two full copies of your project data to remain available for low-latency search. diff --git a/deploy-manage/deploy/elastic-cloud/project-settings.md b/deploy-manage/deploy/elastic-cloud/project-settings.md index 3543bc70d4..26dc1a82b3 100644 --- a/deploy-manage/deploy/elastic-cloud/project-settings.md +++ b/deploy-manage/deploy/elastic-cloud/project-settings.md @@ -54,7 +54,7 @@ $$$elasticsearch-manage-project-search-power-settings$$$ | Setting | Description | Project Type | | :--- | :--- | :--- | | **Search Power** | Search Power controls the amount of baseline resources kept ready for your data and the maximum amount of resources the system can allocate to your project under high search load.

Increase **Search Power** to add resources for improved query throughput and performance as load increases. Decrease it to reduce the amount of provisioned resources, which lowers both active search costs and the reduced rate billed during idle periods, at the cost of more variable query latency. The On-demand setting provides the lowest baseline and idle cost.

Choose from three Search Power settings:

**On-demand:** Low cost option. Query latency is variable and maximum throughput is limited, particularly during demand spikes.

**Performant:** Delivers low _average_ query latency with no practical query throughput limits to meet varying demand.

**High availability:** Maintains consistent low query latency across most search workloads, from steady baseline to peak demand.
| {{es}} | -| **Search Power** | Search Power reserves search capacity for your project. Unlike {{es}} projects, capacity is not scaled automatically based on query load.

The default value is **100**, which reserves enough search resources so that 100% of your project data remains available for low-latency search. You can increase Search Power in increments of 100. Higher values reserve proportionally more capacity: for example, **200** reserves about twice the capacity of **100**.
| {{es}} Vector Database | +| **Search Power** | Search Power reserves search capacity for your project. Unlike {{es}} projects, capacity is not scaled automatically based on query load.

The default value is **100**, which reserves enough search resources so that 100% of your project data remains available for low-latency search. You can increase Search Power in increments of 100. Higher values reserve proportionally more capacity: for example, **200** reserves about twice the capacity of **100**. You can increase Search Power up to **1900**.
| {{es}} Vector Database | | **Search Boost Window** | Non-time series data is always considered search-ready. The **Search Boost Window** determines the volume of time series project data that will be considered search-ready.

Increasing the window results in a bigger portion of time series project data included in the total search-ready data volume.
| {{es}} | | **Data Retention** | Data retention policies determine how long your project data is retained.
In {{serverless-full}} data retention policies are configured through [data streams](../../../manage-data/lifecycle/data-stream.md) and you can [specify different retention periods](../../../manage-data/lifecycle/data-stream/tutorial-update-existing-data-stream.md#set-lifecycle) for specific data streams in your project.

{{elastic-sec}} has two additional configuration settings that can be configured to manage your data retention.

**Maximum data retention period**

When enabled, this setting determines the maximum length of time that data can be retained in any data streams of this project.

Editing this setting replaces the data retention set for all data streams of the project that have a longer data retention defined. Data older than the new maximum retention period that you set is permanently deleted.

**Default data retention period**

When enabled, this setting determines the default retention period that is automatically applied to all data streams in your project that do not have a custom retention period already set.
|Elasticsearch
Observability
Security | | **Project features** | Controls [feature tiers and add-on options](../../../deploy-manage/deploy/elastic-cloud/project-settings.md#project-features-add-ons) for your {{elastic-sec}} project. | Security | diff --git a/solutions/vector-database.md b/solutions/vector-database.md index 72635a8502..fa34a13453 100644 --- a/solutions/vector-database.md +++ b/solutions/vector-database.md @@ -53,7 +53,7 @@ Both the {{es}} Vector Database and the {{es}} project types support [vector sea Choose Vector Database when embeddings and similarity search are central to the workload. For those projects, generate embeddings with managed models through the [Elastic Inference Service](/explore-analyze/elastic-inference/eis.md) or a third-party inference endpoint; you can't deploy custom ML models for in-cluster inference. -Choose the [{{es}} project type](/solutions/elasticsearch-solution-project.md) when you need general-purpose data storage and search, including mixed lexical, time series, and analytics workloads, {{kib}} search tooling such as [Playground](/solutions/elasticsearch-solution-project/playground.md), [Query Rules UI](/solutions/elasticsearch-solution-project/query-rules-ui.md), and [Agent Builder](/explore-analyze/ai-features/elastic-agent-builder.md), or the ability to run custom models on ML nodes. You might also prefer the {{es}} project type if you are an existing {{es}} or OpenSearch user. +Choose the [{{es}} project type](/solutions/elasticsearch-solution-project.md) when you need general-purpose data storage and search, including mixed lexical, time series, and analytics workloads, {{kib}} search tooling such as [Query Rules UI](/solutions/elasticsearch-solution-project/query-rules-ui.md) and [Agent Builder](/explore-analyze/ai-features/elastic-agent-builder.md), or the ability to run custom models on ML nodes. You might also prefer the {{es}} project type if you are an existing {{es}} or OpenSearch user. | Use case | Fit | Why | | --- | --- | --- | From c2a941d2af92b9ee5558d85a34fd9d20b42d3021 Mon Sep 17 00:00:00 2001 From: Vlada Chirmicci Date: Tue, 4 Aug 2026 16:11:04 +0100 Subject: [PATCH 10/13] Add links to the Vector Database project type from the Serverless docs Also update the trial (Sign up) page to include Vector Database --- contribute-docs/how-to/cumulative-docs/guidelines.md | 2 +- deploy-manage/deploy/elastic-cloud/create-an-organization.md | 3 ++- deploy-manage/deploy/elastic-cloud/serverless.md | 5 +++-- 3 files changed, 6 insertions(+), 4 deletions(-) diff --git a/contribute-docs/how-to/cumulative-docs/guidelines.md b/contribute-docs/how-to/cumulative-docs/guidelines.md index 679f3f7a2a..3ea521fb10 100644 --- a/contribute-docs/how-to/cumulative-docs/guidelines.md +++ b/contribute-docs/how-to/cumulative-docs/guidelines.md @@ -304,7 +304,7 @@ For example, a page describing how to create an {{ech}} deployment just requires This is true for most situations. However, it can still be useful to call it out in a few specific scenarios: -* **When there is a high risk of confusion for users**. For example, if a feature is available in two out of three serverless project types, it might make sense to clarify and be explicit about the feature being “unavailable” for the third type. +* **When there is a high risk of confusion for users**. For example, if a feature is available in two out of four serverless project types, it might make sense to clarify and be explicit about the feature being “unavailable” for the other types. ```yml --- diff --git a/deploy-manage/deploy/elastic-cloud/create-an-organization.md b/deploy-manage/deploy/elastic-cloud/create-an-organization.md index b1835c3a67..30e13d5f25 100644 --- a/deploy-manage/deploy/elastic-cloud/create-an-organization.md +++ b/deploy-manage/deploy/elastic-cloud/create-an-organization.md @@ -34,11 +34,12 @@ A deployment lets you explore Elastic solutions for Search, Observability, and S For more information, check the [{{ech}} documentation](cloud-hosted.md). -**Three {{serverless-short}} projects** +**Four {{serverless-short}} project types** {{serverless-short}} projects package {{stack}} features by type of solution: * [{{es}}](../../../solutions/search.md) +* [{{es}} Vector Database](/solutions/vector-database.md) * [Observability](../../../solutions/observability.md) * [Security](../../../solutions/security.md) diff --git a/deploy-manage/deploy/elastic-cloud/serverless.md b/deploy-manage/deploy/elastic-cloud/serverless.md index d87953a845..5713607187 100644 --- a/deploy-manage/deploy/elastic-cloud/serverless.md +++ b/deploy-manage/deploy/elastic-cloud/serverless.md @@ -25,9 +25,10 @@ There are differences between {{es-serverless}} and {{ech}}. Learn more in [Comp ## Get started -Elastic provides three serverless solutions available on {{ecloud}}. Follow these guides to get started with your serverless project: +Elastic provides four serverless project types available on {{ecloud}}. Follow these guides to get started with your serverless project: -* **[{{es-serverless}}](/solutions/search/get-started.md)**: Build powerful applications and search experiences using a rich ecosystem of vector search capabilities, APIs, and libraries. +* **[{{es-serverless}}](/solutions/search/get-started.md)**: Build powerful, scalable search and analytics applications across structured data, logs, metrics, documents, and vectors as part of a broader {{stack}}. +* **[{{es}} Vector Database](/solutions/vector-database/get-started.md)**: Build embedding-driven workloads such as semantic search, RAG, and AI-powered retrieval. Built-in models, `semantic_text`, and vector-optimized defaults mean less configuration and faster time to production. * **[{{obs-serverless}}](../../../solutions/observability/get-started.md)**: Monitor your own platforms and services using powerful machine learning and analytics tools with your logs, metrics, traces, and APM data. * **[{{sec-serverless}}](../../../solutions/security/get-started.md#create-sec-serverless-project)**: Detect, investigate, and respond to threats with SIEM, endpoint protection, and AI-powered analytics capabilities. From 3c74f86d6966644665a71e5dca3ee36a95058bd9 Mon Sep 17 00:00:00 2001 From: Vlada Chirmicci Date: Wed, 5 Aug 2026 13:13:56 +0100 Subject: [PATCH 11/13] Minor tweaks to add the Vector Database project mentioned on CPS and billing pages --- deploy-manage/_snippets/cps-billing.md | 2 +- deploy-manage/cross-project-search-config.md | 2 +- .../differences-from-other-elasticsearch-offerings.md | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-) diff --git a/deploy-manage/_snippets/cps-billing.md b/deploy-manage/_snippets/cps-billing.md index dc7a8a5769..507fcaf8cb 100644 --- a/deploy-manage/_snippets/cps-billing.md +++ b/deploy-manage/_snippets/cps-billing.md @@ -4,7 +4,7 @@ When {{cps-init}} becomes generally available, charges are billed to the origin Usage is metered and charged based on the following factors: -* **Retained data volume:** The volume of data retained in your linked projects each month. The rate varies by the type of each linked project, such as Observability, Security, or {{es}}. For Observability linked projects, the rate also depends on the type of retained data. +* **Retained data volume:** The volume of data retained in your linked projects each month. The rate varies by the type of each linked project, such as Observability, Security, {{es}}, or {{es}} Vector Database. For Observability linked projects, the rate also depends on the type of retained data. * **{{cps-init}} queries:** Cross-project search queries incur a separate data egress charge for data transferred as a result of those queries. This appears as a separate line item on your bill. From August 1, 2026, the [Elastic Cloud Pricing](https://cloud.elastic.co/cloud-pricing-table) page shows detailed pricing for various {{cps}} scenarios. \ No newline at end of file diff --git a/deploy-manage/cross-project-search-config.md b/deploy-manage/cross-project-search-config.md index 0179f154f7..542fe11c3f 100644 --- a/deploy-manage/cross-project-search-config.md +++ b/deploy-manage/cross-project-search-config.md @@ -58,7 +58,7 @@ During technical preview, only newly created projects can be origin projects for To be available for linking, projects must meet the following requirements: - The origin project and all linked projects must be in the same {{ecloud}} organization. -- You can link any combination of {{product.elasticsearch}}, {{product.observability}}, and {{product.security}} projects in the same organization. +- You can link any combination of {{product.elasticsearch}}, {{es}} Vector Database, {{product.observability}}, and {{product.security}} projects in the same organization. - Projects can be linked across cloud providers and regions. For example, a project in GCP `us-east4` can be linked to a project in AWS `eu-central-1` without any additional configuration. - {{sec-serverless}} and {{obs-serverless}} projects require the **Complete** feature tier. Projects on the **Essentials** tier are not compatible with {{cps}}. diff --git a/deploy-manage/deploy/elastic-cloud/differences-from-other-elasticsearch-offerings.md b/deploy-manage/deploy/elastic-cloud/differences-from-other-elasticsearch-offerings.md index 225f5e74b1..c2213fc212 100644 --- a/deploy-manage/deploy/elastic-cloud/differences-from-other-elasticsearch-offerings.md +++ b/deploy-manage/deploy/elastic-cloud/differences-from-other-elasticsearch-offerings.md @@ -25,7 +25,7 @@ The following information reflects our strategic goals, plans and objectives and | **Functionality** | {{ech}} | {{serverless-full}} | |--------|----------------------|--------------------------| | **Management model** | Self-service infrastructure | Fully managed service | -| **Project organization** | Single deployments with multiple capabilities | Separate projects for Elasticsearch, Observability, and Security | +| **Project organization** | Single deployments with multiple capabilities | Separate projects for Elasticsearch, Elasticsearch Vector Database, Observability, and Security | | **Scaling** | Manual or automated with configuration | Fully automated | | **Infrastructure decisions** | User manages capacity | Automatically managed by Elastic | | **Pricing model** | Based on provisioned resources | Based on usage | From a65f69b8d43cee31c05d41973040294d598a71f2 Mon Sep 17 00:00:00 2001 From: Vlada Chirmicci Date: Fri, 14 Aug 2026 06:58:53 +0100 Subject: [PATCH 12/13] increments of 100 statement --- .../billing/vector-database-billing-dimensions.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md b/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md index 2d86d9a76a..266d16880b 100644 --- a/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md +++ b/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md @@ -51,7 +51,7 @@ Vector Database costs follow your storage footprint, search resource allocation, ### Search Power setting [vector-database-billing-search-power-setting] -[Search Power](/deploy-manage/deploy/elastic-cloud/project-settings.md#elasticsearch-manage-project-search-power-settings) reserves search capacity for your project. Start at the default of **100**, measure latency and throughput for your workload, then increase Search Power if you need more capacity. Higher Search Power increases search charges in proportion to the setting and your stored data volume. You can increase Search Power up to **1900**. +[Search Power](/deploy-manage/deploy/elastic-cloud/project-settings.md#elasticsearch-manage-project-search-power-settings) reserves search capacity for your project. Start at the default of **100**, measure latency and throughput for your workload, then increase Search Power if you need more capacity. You increase Search Power in increments of 100. Higher Search Power increases search charges in proportion to the setting and your stored data volume. You can increase Search Power up to **1900**. For high availability, you can increase Search Power to **200**. That setting reserves enough capacity for two full copies of your project data to remain available for low-latency search. From 531826d65929db988e8f962826e17e65eb980717 Mon Sep 17 00:00:00 2001 From: Vlada Chirmicci Date: Fri, 14 Aug 2026 12:02:39 +0100 Subject: [PATCH 13/13] Add the correct product ID related to https://github.com/elastic/docs-content-internal/issues/1534 --- .../billing/vector-database-billing-dimensions.md | 2 +- solutions/vector-database.md | 2 +- solutions/vector-database/get-started.md | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-) diff --git a/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md b/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md index 266d16880b..b74619b690 100644 --- a/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md +++ b/deploy-manage/cloud-organization/billing/vector-database-billing-dimensions.md @@ -4,7 +4,7 @@ applies_to: serverless: ga products: - id: cloud-serverless - - id: elasticsearch + - id: serverless-vector-database description: >- Learn how costs for Elasticsearch Vector Database Serverless projects are calculated across storage, search, ingest, and infrastructure. diff --git a/solutions/vector-database.md b/solutions/vector-database.md index fa34a13453..2d2353a0e1 100644 --- a/solutions/vector-database.md +++ b/solutions/vector-database.md @@ -7,8 +7,8 @@ description: >- optimized for vector workloads, with vector-tuned defaults, hardware profile, inference access, and pricing. It supports semantic and hybrid search. products: - - id: elasticsearch - id: cloud-serverless + - id: serverless-vector-database --- # {{es}} Vector Database project overview diff --git a/solutions/vector-database/get-started.md b/solutions/vector-database/get-started.md index 19e8ab1d5e..b6c7533243 100644 --- a/solutions/vector-database/get-started.md +++ b/solutions/vector-database/get-started.md @@ -6,8 +6,8 @@ description: >- applies_to: serverless: ga products: - - id: elasticsearch - id: cloud-serverless + - id: serverless-vector-database --- # Get started with the {{es}} Vector Database project type