Amazon DocumentDB

Amazon DocumentDB (with MongoDB compatibility) is a document database that AWS runs for you. It stores JSON-like documents and speaks the MongoDB API, so existing MongoDB drivers, frameworks and tools can connect to it, and AWS says applications can typically migrate without code changes or downtime. The service handles patching, backups and monitoring. Compute and storage are separate: a cluster has one primary instance and optional replicas that share a storage volume.

You can choose provisioned instances or DocumentDB Serverless, which scales capacity in fine-grained steps, and a second flavor called Elastic Clusters for very large write and read volumes. Global Clusters replicate data across up to 10 AWS Regions. The product page lists content management, e-commerce, mobile and web apps, and generative AI applications as typical uses.

• MongoDB API compatibility. The service is compatible with MongoDB APIs and drivers. AWS pitches this as the reason a team can move an existing application with few changes, while keeping the tools and frameworks it already uses.
• DocumentDB Serverless. Capacity is measured in DocumentDB Capacity Units (DCUs), billed per second, where one DCU has about 2 GiB of memory with matching CPU and networking. AWS claims up to 90% savings against provisioning for peak capacity.
• Elastic Clusters. A separate cluster type meant to handle millions of reads and writes with petabytes of storage. It is billed on vCPU, database storage and backup storage, rather than on instances and I/O.
• Global Clusters. Optional cross-region replication with latency below one second, using dedicated infrastructure. It supports recovery from a region-wide outage and lets applications read from the nearest cluster. Replicated write I/O is charged.
• Two storage configurations. Standard bills I/O per million requests, while I/O-Optimized includes I/O in the price. AWS suggests I/O-Optimized when I/O would exceed 25% of the cluster spend, and Standard when it stays under that.
• Pause and cost controls. Compute instances can be stopped for up to seven days, which suits a test cluster over a weekend. Backup storage equal to 100 percent of your data storage is free each month, and per-second billing applies to instances.

• A team running MongoDB on its own servers. Wants to stop patching, backing up and monitoring the database by hand. The API compatibility lets the application keep its drivers, and the managed service takes over maintenance, though the team should test every query pattern it relies on before cutover.
• A SaaS product with spiky, hard-to-predict traffic. Serverless capacity follows demand in DCU steps, which avoids paying for peak size all day. The vendor example scales from 0.5 to 7 DCUs within minutes. Read replicas and Multi-AZ deployment remain available.
• A retailer storing profiles, reviews and catalog content. E-commerce and content management are the vendor's own examples. Document storage suits records that vary in shape, and Global Clusters add low-latency reads for shoppers in other regions, at the price of replicated write I/O.

You need an AWS account and create a cluster in the console, choosing provisioned instances, Serverless or Elastic Clusters, and one of the two storage configurations. Applications then connect with a MongoDB driver. AWS says migration is typically possible without code changes or downtime, and the product page links hands-on training, documentation and a console entry.

A pricing calculator estimates cost before you commit. A development cluster can run on a single instance, because instances do not hold data, and the pricing page calls this a popular approach for development. Learning time depends on prior MongoDB and AWS networking knowledge.

DocumentDB supports MongoDB 3.6, 4.0, 5.0 and 8.0 drivers and says a vast majority of applications work. Instances bill per second with a 10-minute minimum. Serverless, Elastic Clusters and Global Clusters are listed.

In our database management systems ranking Amazon DocumentDB stands at #17 of 22: the products next to it are Google Firestore above and Azure Cosmos DB below. No numeric score is attached; established products follow the challengers in their ranked order.

North America is home to 15 of the 22 vendors, Amazon DocumentDB among them (the United States).

The I/O-Optimized configuration removes the I/O charge and AWS suggests it when I/O exceeds 25 percent of spend.

Instances can be paused up to seven days.

In the catalogue: Instance seconds, I/O, storage and backup billed separately; I/O-Optimized option; one-month free trial.

AWS prices DocumentDB on four dimensions: on-demand instances billed per second with a 10-minute minimum, database I/O per million requests, database storage per GB-month, and backup storage above your free allowance. I/O-Optimized drops the I/O dimension. Serverless is billed per DCU-hour; the page works an example at $0.0905 per DCU-hour. Elastic Clusters bill vCPU, storage and backup. A free trial inside the AWS Free Tier gives 750 hours of db.t3.medium, 30 million IOs, 5 GB storage and 5 GB backup for 30 days. Per-instance prices vary by region and were loaded dynamically, so they are not quoted here. Database Savings Plans offer discounts for a one-year commitment.

Where this comes from: the vendor pages at aws.amazon.com/documentdb, checked for Best Database Management Systems in 2026 on 7 November 2026.

AWS describes security and compliance built for high-sensitivity organizations in banking, healthcare and the public sector, but the product and pricing pages checked do not list individual certifications. Data lives in the AWS Region you pick, and the free trial is unavailable in AWS GovCloud (US) and China (Ningxia).

Integrations with AWS IAM, AWS Backup and CloudWatch are named. Because the API follows MongoDB, data can be read out through standard MongoDB tooling, though AWS does not describe an export procedure on these pages.

Integrations the vendor names (7):

• Amazon CloudWatch
• AWS IAM
• AWS Backup
• Amazon OpenSearch Service (zero-ETL)
• MongoDB drivers and tools
• AWS Pricing Calculator
• Database Savings Plans

These are the nearest neighbours of Amazon DocumentDB in the same list, picked on shared traits and price model.

• Oracle Database (#11 of 22): Commercial enterprise database, now branded Oracle AI Database, with capped free edition. Best for: Large organisations already tied to Oracle licensing, tooling and staff.
• Amazon DynamoDB (#12 of 22): Best for: AWS-based teams wanting a database with no servers to patch. (United States)
• Azure Cosmos DB (#18 of 22): Aimed at: Azure teams wanting turnkey multi-region distribution.

Read from the vendor’s own pages on 5 November 2026.

• aws.amazon.com/documentdb: product description, benefits, use cases, Global Clusters, integrations
• aws.amazon.com/documentdb/pricing: pricing dimensions, storage configurations, serverless DCU, free trial, pause and backup details

We rank on what separates products after the demo: time to useful, the real price, the exit and who the product is for. That puts Amazon DocumentDB at #17 of 22. The editorial process sets out how facts are gathered and how commercial links are handled.