Overview

Amazon Keyspaces is a managed, serverless database service compatible with Apache Cassandra. It is best suited to teams building or migrating applications that already use Cassandra’s query API and tooling. Its strongest case is managed scale and regional deployment without infrastructure management; capacity-based charges and a three-month free-tier window make cost planning important.

Key features

Keyspaces supports the Cassandra Query Language API, existing application code, drivers, and developer tools. That compatibility can ease migration and let teams retain familiar interfaces, but it is most useful when Cassandra compatibility is a requirement rather than an incidental preference.

Teams can choose on-demand capacity, which charges for reads and writes performed and adjusts to traffic, or provisioned capacity, where expected read and write capacity is specified in advance. Provisioned capacity can use auto scaling. On-demand suits variable traffic when avoiding capacity planning matters; provisioned capacity gives teams more control over capacity settings, though they pay for provisioned capacity as well as storage and other applicable usage.

AWS describes single-digit-millisecond read and write performance, and tables have no limit on size or row count. The maximum row size is 1 MB. For workloads that need geographic deployment, Multi-Region deployments offer up to 99.999% availability; table data is also replicated three times across multiple Availability Zones. These are useful resilience characteristics, but the stated availability figure is specifically for Multi-Region deployments.

Point-in-time recovery uses continuous backups and can restore table data to any second in the preceding 35 days. Managed Time to Live expires rows and attributes without requiring users to manage tombstones or compaction. Keyspaces Streams provides change data capture through its Streams API or the Kinesis Client Library, while changes can also stream to Amazon OpenSearch Service. CloudWatch integration provides performance metrics.

Security includes encryption at rest by default, with AWS-owned or customer-managed keys through AWS Key Management Service. AWS Identity and Access Management integration, TLS, and AWS PrivateLink connectivity to Amazon VPC provide access and network controls. Keyspaces supports tunable consistency and transactions, giving applications controls beyond its Cassandra-compatible query interface.

Pricing

Keyspaces is paid, with two capacity modes rather than a published monthly price for either mode. On-demand capacity charges for reads and writes performed, plus storage and other applicable usage; it fits traffic that changes enough to make fixed capacity less attractive. Provisioned capacity charges for capacity specified in advance, storage, and other applicable usage; it fits teams able to estimate throughput and use auto scaling to adjust capacity. Neither mode has a listed price, so expected read, write, and storage use should guide the cost decision.

The AWS Free Tier costs 0.00 USD per free and applies for the first three months after creating the first Keyspaces resource. It includes 30 million write request units, 30 million read request units, and 1 GB of storage, and is limited to one free tier per payer account. This is a useful initial allowance, not an ongoing free plan.

Database Savings Plans require a usage commitment over a 1-year term, with up to 18% savings stated by AWS. The plan has no listed price, so its appeal depends on whether a team can commit to usage that justifies the stated potential savings. On-demand and provisioned capacity descriptions also state a monthly allowance of 30 million write request units, 30 million read request units, and 1 GB of storage free for the first three months.

Platforms

Keyspaces is a managed service available through API and web platforms. Its query interface is CQL, and it supports horizontal scaling and multi-region clusters.

Who it's for

Keyspaces fits teams migrating Cassandra workloads or building low-latency applications, IoT device data stores, game player profiles, and time-series workloads.

Pros and cons

Pro: Cassandra API, code, drivers, and tools support migration without requiring a new query interface.
Pro: On-demand and provisioned capacity offer different ways to match changing or anticipated traffic, with auto scaling available for provisioned capacity.
Pro: Point-in-time recovery reaches any second in the previous 35 days, and TTL avoids manual tombstone and compaction management.
Pro: Multi-Region deployments offer up to 99.999% availability, with triple replication across Availability Zones.
Con: The free tier lasts only three months and is restricted to one per payer account, so it does not provide continuing free usage.
Con: Capacity, storage, and applicable usage charges have no listed prices, making costs harder to compare before estimating workload demand.
Con: Maximum row size is 1 MB, a constraint for applications that need larger rows.

Alternatives

Wide-Column Databases and Database Software are useful starting points for comparing options across those categories.

Choose Google Cloud Bigtable if its paid plans and free trial suit your evaluation; its Enterprise plan is 0.65 USD per month and Enterprise Plus is 0.85 USD per month, while Data Boost is 0.00 USD per month.

Choose Apache HBase if an open-source database that operators configure and manage better fits your deployment. It is 0.00 USD per free and supports API, Linux, and self-hosted platforms.

SWC-DB is a free, self-hosted alternative with GPLv3 licensing; commercial-purpose licensing inquiries apply to uses limited by GPLv3.

Cloudera Data Lake Service is a paid alternative with custom pricing through sales and no free plan.

Elassandra is a free alternative for API, Linux, and self-hosted use.

Baidu Tera is a free option with a free plan and API, Linux, and self-hosted platforms.

HPE Ezmeral Data Fabric offers a free plan and trial across API, Linux, self-hosted, and web platforms; its software pricing varies by reseller.

ScyllaDB Manager is a free option for API, Linux, and self-hosted use, with enterprise clusters and custom licensing available by quote.

Verdict

Choose Amazon Keyspaces when Cassandra compatibility and managed scaling matter more than having a published capacity price. Its recovery, access controls, and multi-region option make it a capable fit for Cassandra-oriented workloads; look elsewhere if ongoing free use or clear upfront pricing is essential.

	Amazon Keyspaces	Google Cloud Bigtable
Free plan	Yes	No
Free trial	No	Yes
Paid from	Free	$0.03/mo
Open source	No	No
Platforms	api, Web	api, Web
Free plan	No	No

Amazon Keyspaces vs Google Cloud Bigtable

Amazon Keyspaces

Google Cloud Bigtable