Managed services - verified October 18, 2026

Elasticsearch (vector search)

Elasticsearch is listed under managed services (deployment type: managed serverless). First listed price, in its own unit: As low as $0.14, per VCU-hour ingest (Serverless). Max dimensions: 4,096. Stated: hybrid and sparse search and built-in embedding. Read from the vendor's own pages on October 18, 2026.

Read from elastic.co, raw.githubusercontent.com. Docs - Pricing - License - Compare every vector database

Status

Availability label: Not stated
No GA or preview label for dense_vector search is quoted from the pages checked.

How Elasticsearch charges

Every price Elasticsearch states, in its own unit, with the quote and where it comes from. Nothing is converted.

4 price rows in each product's own unit, grouped by unit and sorted only inside a unit
Price	Unit, as stated	Quote and source	Where it comes from
Per VCU-hour
As low as $0.14	per VCU-hour ingest (Serverless)	“Ingest*Per VCU per hour As low as $0.14” elastic.co	-
As low as $0.09	per VCU-hour search (Serverless)	“Search*Per VCU per hour As low as $0.09” elastic.co	-
As low as $0.07	per VCU-hour machine learning (Serverless)	“Machine LearningPer VCU per hour As low as $0.07” elastic.co	-
Per GB-month stored
As low as $0.047	per GB-month retained (Serverless)	“Storage & RetentionPer GB retained per month As low as $0.047” elastic.co	-

Free tier and limits

Free tier: 14-day free trial
“New users can sign up for a free 14-day trial.”
elastic.co
“Vector profiles receive 50 GB free”
elastic.co
Limits: Not stated
No vector-count limit is stated on the pages checked.

Search and indexing

Max dimensions: 4,096
“Number of vector dimensions. Can't exceed 4096.”
elastic.co
Distance metrics: l2_norm, dot_product, cosine, max_inner_product
“Valid values for similarity l2_norm Computes similarity based on the L2 distance”
elastic.co
“max_inner_product Computes the maximum inner product of two vectors.”
elastic.co
Index types: hnsw, int8 / int4 quantized variants, bbq_hnsw, bbq_flat, bbq_disk
“The supported vector quantization strategies for dense_vector kNN indexing are: - int8 - int4 - bbq, available as: - bbq_hnsw - bbq_flat - bbq_disk”
elastic.co
“the default index type is hnsw with no quantization.”
elastic.co
Hybrid and sparse search: Yes: full-text plus vector search, RRF recommended
“Hybrid search runs full-text search and vector search in one request.”
elastic.co
“We recommend implementing hybrid search with the reciprocal rank fusion (RRF) algorithm.”
elastic.co
Metadata filtering: Not stated
Not described on the dense_vector, semantic_text or hybrid pages checked.
Multi-tenancy: Not stated
Not described on the pages checked.

Built-in embedding and reranking

Built-in embedding: Yes, via semantic_text inference fields
“Generates embeddings during indexing: Automatically generates embeddings when you index documents”
elastic.co
Built-in reranking: Not stated
Not stated on the pages checked.

Deployment, license and SDKs

Regions and clouds: Not stated
Not stated on the pages checked.
BYOC and self-host: Source code available under AGPL-3.0, SSPL-1.0 or Elastic License 2.0
“Source code in this repository is covered by (i) a triple license under the "GNU Affero General Public License v3.0 only", "the Server Side Public License, v 1", and the "Elastic License 2.0"”
raw.githubusercontent.com
License: AGPL-3.0-only, SSPL-1.0 or Elastic License 2.0 (triple license by default)
“Source code in this repository is covered by (i) a triple license under the "GNU Affero General Public License v3.0 only", "the Server Side Public License, v 1", and the "Elastic License 2.0"”
raw.githubusercontent.com
SDKs: Python, Java, JavaScript/Node.js, C#/.NET, PHP, Go, Ruby, Rust (experimental)
“Python pip install elasticsearch docs repo Java ( project configuration) docs repo JavaScript / Node.js npm install @elastic/elasticsearch”
elastic.co
“C# / .NET dotnet add package Elastic.Clients.Elasticsearch docs repo PHP composer require elasticsearch/elasticsearch docs repo Go require github.com/elastic/go-elasticsearch/v9 9.0 docs repo Ruby gem install elasticsearch docs repo Rust (experimental) cargo add elasticsearch”
elastic.co

Vendor-stated claims

Performance, scale and SLA figures as Elasticsearch states them. Not measured.

BBQ reduces memory footprint by 96% (32x), at a cost of accuracyvendor-stated

“This reduces the memory footprint by 96% (or 32x) at a larger cost of accuracy.”

elastic.co

Research notes

Pricing shown is Elastic Cloud Serverless (Search) "as low as" rates; the hybrid search docs recommend an "Elasticsearch Vector Database project" on Serverless. In Elastic Stack 9.0 float dense vectors are always indexed as int8_hnsw per the docs.

FAQ

How much does Elasticsearch cost?

In Elasticsearch's own units: As low as $0.14 (per VCU-hour ingest (Serverless)); As low as $0.09 (per VCU-hour search (Serverless)); As low as $0.07 (per VCU-hour machine learning (Serverless)); As low as $0.047 (per GB-month retained (Serverless)). As of October 18, 2026; prices are not converted.

What is the maximum vector dimension in Elasticsearch?

4,096, per the vendor: "Number of vector dimensions. Can't exceed 4096."

Does Elasticsearch support hybrid search?

Yes: full-text plus vector search, RRF recommended, per the vendor: "Hybrid search runs full-text search and vector search in one request."

Does Elasticsearch have built-in embedding or reranking?

Embedding: Yes, via semantic_text inference fields. Reranking: not stated. Not stated on the pages checked.

Sources

Last verified 2026-10-18 from each vendor's own docs, pricing pages, READMEs and license files; every quote was string-matched against the page that day. This is a dated snapshot, not a benchmark: prices are shown in each product's own unit and change without notice, so check the linked source before you commit.

Newsletter

One email per week. New model releases, retrieval tooling and pricing changes and what they mean for what you build.