Cloud provider services - verified
Amazon OpenSearch Serverless (vector search collections)
Amazon OpenSearch Serverless is listed under cloud provider services (deployment type: cloud-native service). The pricing page describes OCU-hour billing ("One OCU comprises approximately 6 GB of RAM and corresponding vCPU"), but the per-OCU rates appear only in its pricing calculator, not in the page text. Per-OCU compute rates appear only in a pricing calculator. Max dimensions: 16,000. Stated: hybrid and sparse search. Read from the vendor's own pages on October 18, 2026.
Read from docs.aws.amazon.com, aws.amazon.com. Docs - Pricing - Compare every vector database
Status
- Availability label
- Not stated
No GA or preview label is quoted from the pages checked.
- NextGen and Classic
Amazon OpenSearch Serverless splits vector collections into NextGen and Classic
AWS now distinguishes NextGen and Classic vector search collections. The pricing page bills NextGen collections for compute and storage separately; the per-OCU rate itself appears only in the pricing calculator.
“With Amazon OpenSearch Serverless NextGen collections, you only pay for the resources consumed by the workload. OpenSearch Serverless charges for compute and storage separately.”
aws.amazon.com “With Amazon OpenSearch Serverless Classic collections, you only pay for the resources consumed by the workload.”
aws.amazon.com Research note: AWS now distinguishes NextGen and Classic vector search collections; NextGen collections have a 10-second read-after-write refresh interval per the docs. Source page
How Amazon OpenSearch Serverless charges
Every price Amazon OpenSearch Serverless states, in its own unit, with the quote and where it comes from. Nothing is converted.
| Product | Rate | Why | Billing model, quoted |
|---|---|---|---|
| Amazon OpenSearch Serverless | Not statedPer-OCU compute rates | The pricing page describes OCU-hour billing ("One OCU comprises approximately 6 GB of RAM and corresponding vCPU"), but the per-OCU rates appear only in its pricing calculator, not in the page text. | Compute (OCUs) and storage billed separately “With Amazon OpenSearch Serverless NextGen collections, you only pay for the resources consumed by the workload. OpenSearch Serverless charges for compute and storage separately.” |
The pricing page describes OCU-hour billing ("One OCU comprises approximately 6 GB of RAM and corresponding vCPU"), but the per-OCU rates appear only in its pricing calculator, not in the page text.
Free tier and limits
- Free tier
- Not stated
Not stated on the pricing page checked.
- Limits
- Not stated
Not stated on the pages checked.
Search and indexing
- Max dimensions
16,000
“The vector engine provides distance metrics such as Euclidean distance, cosine similarity, and dot product similarity, and can accommodate 16,000 dimensions.”
docs.aws.amazon.com - Distance metrics
Euclidean, cosine, dot product
“The vector engine provides distance metrics such as Euclidean distance, cosine similarity, and dot product similarity, and can accommodate 16,000 dimensions.”
docs.aws.amazon.com - Index types
- Not stated
The collection docs say the engine uses the OpenSearch k-NN feature but do not list index methods for serverless collections.
- Hybrid and sparse search
Full-text search alongside vector search in the same collection
“With these operations, you can use full-text search, advanced filtering, aggregations, geospatial queries, and nested queries”
docs.aws.amazon.com - Metadata filtering
Yes
“You can store fields with various data types for metadata, such as numbers, Booleans, dates, keywords, and geopoints.”
docs.aws.amazon.com - Multi-tenancy
- Not stated
Not described on the pages checked.
Built-in embedding and reranking
- Built-in embedding
- Not stated
Not stated on the pages checked.
- Built-in reranking
- Not stated
Not stated on the pages checked.
Deployment, license and SDKs
- Regions and clouds
- Not stated
AWS only; specific regions are not listed on the pages checked.
- BYOC and self-host
- Not stated
Managed AWS service; self-hosting is not applicable (see the OpenSearch open-source entry).
- License
- Not stated
Managed service; no license stated.
- SDKs
- Not stated
Not listed on the pages checked.
Vendor-stated claims
Performance, scale and SLA figures as Amazon OpenSearch Serverless states them. Not measured.
NextGen collections create all indexes with 32x compressionvendor-statedinferred
Our reading: The quote itself does not mention NextGen collections.
“All indexes are created with advanced 32x compression technique by default.”
Research notes
AWS now distinguishes NextGen and Classic vector search collections; NextGen collections have a 10-second read-after-write refresh interval per the docs. The pricing page also says: "With Amazon OpenSearch Serverless Classic collections, you only pay for the resources consumed by the workload."
FAQ
How much does Amazon OpenSearch Serverless cost?
Per-OCU compute rates: not stated. The pricing page describes OCU-hour billing ("One OCU comprises approximately 6 GB of RAM and corresponding vCPU"), but the per-OCU rates appear only in its pricing calculator, not in the page text.
What is the maximum vector dimension in Amazon OpenSearch Serverless?
16,000, per the vendor: "The vector engine provides distance metrics such as Euclidean distance, cosine similarity, and dot product similarity, and can accommodate 16,000 dimensions."
Does Amazon OpenSearch Serverless support hybrid search?
Full-text search alongside vector search in the same collection, per the vendor: "With these operations, you can use full-text search, advanced filtering, aggregations, geospatial queries, and nested queries"
Does Amazon OpenSearch Serverless have built-in embedding or reranking?
Embedding: not stated. Not stated on the pages checked. Reranking: not stated. Not stated on the pages checked.
Sources
Last verified 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
Know when a vector database changes its pricing
One email per week. New model releases, retrieval tooling and pricing changes and what they mean for what you build.