Services or capabilities described in Amazon Web Services documentation might vary by Region. To see the differences applicable to the China Regions, see Getting Started with Amazon Web Services in China (PDF).

On-demand capacity mode

Amazon DynamoDB on-demand is a serverless billing option that can serve millions of requests per second without capacity planning. DynamoDB on-demand offers pay-per-request pricing for read and write requests so that you only pay for what you use.

When you choose on-demand mode, DynamoDB instantly accommodates your workloads as they ramp up or down to any previously reached traffic level. If a workload’s traffic level hits a new peak, DynamoDB adapts rapidly to accommodate the workload. For more information about on-demand mode's scaling properties, see Initial throughput and scaling properties.

Tables that use on-demand mode deliver the same single-digit millisecond latency, service-level agreement (SLA) commitment, and security that DynamoDB already offers. You can choose on-demand for both new and existing tables and you can continue using the existing DynamoDB APIs without changing code.

The on-demand throughput rate is bounded by the table-level throughput quota which applies to all tables with the account. You can request an increase for this quota. For more information, see Throughput default quotas.

Optionally, you can also configure maximum read or write (or both) throughput per second for individual on-demand tables and global secondary indexes. By configuring throughput, you can keep table-level usage and costs bounded, protect against inadvertent surge in consumed resources, and prevent excessive use for predictable cost management. Throughput requests that exceed the maximum table throughput are throttled. You can modify the table-specific maximum throughput at any time based on your application requirements. For more information, see Maximum throughput for on-demand tables.

To get started, create or update a table to use on-demand mode. For more information, see Basic operations on DynamoDB tables.

You can switch tables from on-demand mode to provisioned capacity mode at any time. When you do multiple switches between capacity modes, the following conditions apply:

For more information about switching between read and write capacity modes, see Considerations when switching capacity modes.

Read request units and write request units

DynamoDB charges you for the reads and writes that your application performs on your tables in terms of read request units and write request units.

One read request unit represents one strongly consistent read operation per second, or two eventually consistent read operations per second, for an item up to 4 KB in size. For more information about DynamoDB read consistency models, see Read consistency.

One write request unit represents one write operation per second, for an item up to 1 KB in size.

For more information about how read and write units are consumed, see Read and write operations.

Initial throughput and scaling properties

DynamoDB tables using on-demand capacity mode automatically adapt to your application’s traffic volume. New on-demand tables will be able to sustain up to 4,000 writes per second and 12,000 reads per second. On-demand capacity mode instantly accommodates up to double the previous peak traffic on a table. For example, say that your application’s traffic pattern varies between 25,000 and 50,000 strongly consistent reads per second. 50,000 reads per second is the previous traffic peak. On-demand capacity mode instantly accommodates sustained traffic of up to 100,000 reads per second. If your application sustains traffic of 100,000 reads per second, that peak becomes your new previous peak. This previous peak enables subsequent traffic to reach up to 200,000 reads per second.

If your workload generates more than double your previous peak on a table, DynamoDB automatically allocates more capacity as your traffic volume increases. This capacity allocation helps ensure that your workload doesn't experience throttling. However, throttling can occur if you exceed double your previous peak within 30 minutes. For example, say that your application’s traffic pattern varies between 25,000 and 50,000 strongly consistent reads per second. 50,000 reads per second is the previously reached traffic peak. We recommend that you either pre-warm your table or space your traffic growth over at least 30 minutes before driving more than 100,000 reads per second. For more information about pre-warming, see Pre-warming a table for on-demand capacity mode.