Amazon Athena
Database management systems (DBMS)
Database as a service (DBaaS) providers
Database software
- Features
- Ease of use
- Ease of management
- Quality of support
- Affordability
- Market presence
Take the quiz to check if Amazon Athena and its alternatives fit your requirements.
Pay-as-you-go
Free Trial unavailableFree version unavailable
Small
Medium
Large
- Retail and wholesale
- Media and communications
- Agriculture, fishing, and forestry
What is Amazon Athena
Amazon Athena is a serverless, managed query service that runs SQL queries directly against data stored in Amazon S3 and other supported sources. It is primarily used by data analysts, data engineers, and BI teams for ad hoc analysis, log analytics, and querying data lake datasets without provisioning database infrastructure. Athena uses open table formats and engines (such as Trino/Presto and Apache Spark) and integrates with the AWS Glue Data Catalog for schema and metadata management. It is typically positioned as a query layer over object storage rather than a general-purpose transactional database.
Serverless, on-demand querying
Athena does not require users to provision or manage database servers, which reduces operational overhead for ad hoc and intermittent workloads. Capacity scales automatically within AWS-managed limits, making it practical for bursty query patterns. This model fits teams that want SQL access to data lake files without standing up a dedicated warehouse cluster.
Strong S3 data lake fit
Athena queries data in place in Amazon S3, which supports common analytics formats such as Parquet, ORC, JSON, and CSV. It works with the AWS Glue Data Catalog for centralized table definitions and partitions, supporting shared metadata across multiple AWS analytics services. This is useful for organizations standardizing on S3 as the storage layer for analytics and governance.
Multiple engines and connectors
Athena supports different execution options, including a SQL engine based on Trino/Presto and an Apache Spark option for certain workloads. It also supports federated querying via connectors to query some external data sources without moving data into S3 first. This flexibility can reduce the need to duplicate datasets into a separate database for exploration and reporting.
Not built for OLTP
Athena is designed for analytical querying and is not a transactional DBMS for high-concurrency inserts/updates with strict latency requirements. Data changes typically occur by writing new files/partitions to S3 and updating metadata, which differs from row-level transactional behavior. Teams needing frequent small writes or low-latency point lookups often require a different database service.
Performance depends on data layout
Query speed and cost depend heavily on file format, partitioning, and data organization in S3. Poorly partitioned tables, many small files, or wide scans can lead to higher data scanned and longer runtimes. Achieving consistent performance often requires data engineering work (compaction, partition strategy, columnar formats) rather than only SQL tuning.
AWS-centric governance and portability
Athena is tightly integrated with AWS identity, networking, and catalogs, which can increase coupling to AWS for access control and metadata management. While it uses common SQL engines and open formats, operational patterns (catalogs, permissions, workgroups, and cost controls) are AWS-specific. Organizations with multi-cloud requirements may need additional tooling to standardize governance and query access across environments.
Plan & Pricing
Pricing model: Pay-as-you-go SQL (on-demand) queries: $5.00 per TB scanned (example/default rate shown on the Athena pricing page). Billed based on amount of data scanned by the query (example on page uses $5/TB). Federated queries are billed per TB scanned aggregated across data sources, rounded up to the nearest megabyte with a 10 MB minimum per query (unless Provisioned Capacity is used). Provisioned Capacity / Capacity Reservations: $0.30 per DPU-hour (example rate shown for Provisioned Capacity calculations). Apache Spark (notebook / Spark engine): $0.35 per DPU-hour (rate shown for Spark application example). Additional costs / notes: Querying reads data from Amazon S3 (standard S3 storage, request, and data transfer rates apply). Using AWS Glue Data Catalog incurs standard Glue Data Catalog charges. Federated queries may invoke AWS Lambda and therefore incur Lambda charges at standard rates.
Seller details
Amazon Web Services, Inc.
Seattle, Washington, USA
2006
Subsidiary
https://aws.amazon.com/
https://x.com/awscloud
https://www.linkedin.com/company/amazon-web-services/
