Cross-Vendor Document Database Pricing Comparison (DocumentDB Cosmos DB DynamoDB)

Parent: Atlas Cost Planning: Region Pricing, Estimation, Scheduled Scaling and Competitor TCO · Published reference · snapshot 2026-10-19 · skill mongodb-operations-expert/references/atlas-cost-planning-region-pricing-estimation-scheduled-scaling-and-competitor-tco.md

↓ Facts as markdown all context files

Atlas competitors bill on different units: Amazon DocumentDB charges per instance-hour plus storage (Standard per-I/O or I/O-Optimized), Azure Cosmos DB charges per provisioned or consumed Request Unit plus storage in every replicated region, and DynamoDB charges per request unit on-demand or provisioned, with on-demand prices cut 50% from 2024-11-01. The two public Atlas-versus-competitor comparisons are vendor-authored and conflict: Microsoft claims Azure DocumentDB is 39-70% cheaper than Atlas on AWS, while Vantage found Amazon DocumentDB 26.7% dearer than Atlas, partly because Atlas includes replica nodes in its tier price. The pack holds 115 source-anchored units from 6 sources (5 hosts) across 10 facets, mostly vendor pricing-page rates and worked examples.

These notes link each claim to its source. Some citations identify archived research reports that are no longer hosted on this site. A published reference means the content is available; it does not certify independent review or accuracy.Read the editorial policy and follow the sources before relying on a claim.

Definitions

Name clash: 'DocumentDB' means Amazon DocumentDB (AWS, Mongo-compatible, instance plus I/O billing) and, in Microsoft's 2026 material, 'Azure DocumentDB' (vCore Mongo-compatible service in the Cosmos DB family). Cite the vendor prefix in any TCO table. [source]
What you've effectively purchased is a credit of $8 per hour, for 100 K RU/sec using the list price in East US, at the price of $6.40 per hour. You can then draw down from this pre-paid throughput reservation on an hourly basis for the provisioned throughput capacity in any global Azure region at the respective regional list prices set for your subscription. In this example, where you provision 50 K RU/sec each in East US, and Japan East, you will be able to draw $8.00 worth of provisioned throughput per hour, and will be billed the overage of $0.50 per hour (or $360/month). [source]
If all you know is the number of vcores and servers in your existing database cluster, read about estimating request units using vCores or vCPUs [source]
Azure DocumentDB is a fully managed, MongoDB-compatible database on Azure, built on an MIT-licensed open-source engine that runs consistently across on-premises, hybrid, and cloud environments. Developers keep the skills and tooling they already have. Infrastructure teams get a pricing model built on compute and storage, with no licensing fees and no throughput unit calculations. [source] — Azure DocumentDB, not Amazon DocumentDB
Azure DocumentDB is a fully managed enterprise-grade MongoDB-compatible database and vector database for modern app development, including AI applications. With its predictable low costs, Open-source project, as well as 99.03% MongoDB compatibility, it is ideal for any MongoDB application running on Azure. [source] — Azure DocumentDB
Amazon DocumentDB (with MongoDB compatibility) is a fully managed document database service that supports MongoDB workloads. With Amazon DocumentDB, you only pay for what you use, and there are no upfront costs. Amazon DocumentDB pricing has four dimensions: [source]
Amazon DocumentDB Elastic Clusters are a new type of Amazon DocumentDB cluster that lets you elastically scale your document database to handle millions of reads and writes with petabytes of storage capacity. Pay only for what you use, and there are no upfront costs. Amazon DocumentDB Elastic Clusters pricing has three dimensions: [source]

Structure and components

Amazon DocumentDB instance-based clusters bill four dimensions on Standard storage (instance-hours, storage, I/O per million, backup beyond free allotment) and three on I/O-Optimized (no I/O charge). AWS lists storage at $0.10/GB-month and I/O at $0.20 per million on Standard, and storage at $0.30/GB-month on I/O-Optimized. [source] — Amazon DocumentDB billing dimensions
Both provisioned and on-demand billing modes use the same underlying infrastructure to achieve high availability, scale, reliability, and performance. The key differences are that on-demand is always 100% utilized due to pay-per-request billing and on-demand scales transparently, without needing to specify a scaling policy. As a result, many customers prefer the simplicity of on-demand mode to build modern, serverless applications that can start small and scale to millions of requests per second. Continually working backward from our customer feedback, in early 2024, we launched configurable m [source]
If you are new to Amazon DocumentDB, you are eligible for a one month free trial as part of AWS Free Tier. This free trial includes 750 hours per month of db.t3.medium instance usage, 30 million IOs, 5 GB of storage, and 5 GB of backup storage for free for 30 days. After the one-month free trial expires or your usage exceeds the free allowance, you can shut down your cluster to avoid any charges or keep it running at our standard on-demand rates. [source]
Billing is for any portion of a wall-clock hour, not a 60 minute duration. All the examples shown in this doc are based on the price for an Azure Cosmos DB account deployed in a non-government region in the US. The pricing and calculation vary depending on the region you are using, see the Azure Cosmos DB pricing page for latest pricing information. [source]
Amazon DocumentDB was designed to eliminate unnecessary I/O operations in order to reduce costs and to ensure resources are available for serving read/write traffic. Write I/Os are only consumed when pushing write-ahead log records to the storage layer for the purpose of making writes durable. Write I/Os are counted in 4 KB units. [source]

How it works

Cosmos DB throughput and storage are billed in every replicated region: total RU/s billed = RU/s x number of regions. Billing is per wall-clock hour, so a scale-up for part of an hour charges the full hour at the higher rate. Autoscale bills the peak RU/s each hour (10-100% of the set maximum) at a 1.5x single-write-region multiplier. [source] — Azure Cosmos DB regions and autoscale
When we launched DynamoDB in 2012, provisioned capacity was the only throughput option available. Provisioned capacity requires you to predict and plan your throughput requirements. For provisioned tables, you must specify how much read and write throughput per second you require for your application, and you’re charged based on the hourly read and write capacity you have provisioned, not how much your application has consumed. In 2017, we added provisioned auto scaling to help improve scaling and utilization. Although it was effective, we learned that customers wanted a serverless experience [source]
Azure Cosmos DB reserved capacity enables you to purchase provisioned throughput in advance (a reserved capacity or a reservation) that can be applied to all Azure Cosmos DB databases and containers (for any API or data model) across all Azure regions. Because provisioned throughput price varies per region, it helps to think of reserved capacity as a monetary credit that you've purchased at a discount, that can be drawn from for the provisioned throughput at the respective price in each region. For example, let's say you have an Azure Cosmos DB account with a single container provisioned with [source]
While on-demand was previously cost-effective for spiky workloads, with this pricing change, most provisioned capacity workloads on DynamoDB today will achieve a lower price with on-demand mode. This pricing change is transformative because it makes on-demand the default and recommended mode for most DynamoDB workloads. Whether you’re running a new application or a well-established one, on-demand mode simplifies the operational experience, while providing seamless scalability and responsiveness to handle changes to your traffic pattern, making it an ideal solution for startups, growing applica [source]
When using autoscale, the entry point scale range you can set is 100 - 1000 RU/s. If you want to use autoscale and keep your free tier account completely free, create either one container with this scale range, or a shared throughput database with up to 25 containers inside. The example below illustrates how billing works if you provision throughput higher than the 100 - 1000 RU/s scale range. [source]
With Amazon DocumentDB, you have the flexibility to choose your storage configuration for every database cluster you have. In order to properly allocate your clusters between standard and I/O-optimized, you can track your Amazon DocumentDB costs cluster-wise. To do so, you can add tags to existing clusters, enable cost allocation tagging in your [AWS Billing and Cost Management dashboard](https://aws.amazon.com/pricing/), and analyze your costs for a given cluster in the [AWS Cost Explorer Service](https://aws.amazon.com/aws-cost-management/aws-cost-explorer/). For information on cost analysis [source]
Scale to zero throughput cost – With DynamoDB, the throughput a table is a capable of serving at any given moment is decoupled from what you are billed. For example, an on-demand table may be capable of serving 20,000 reads and 20,000 writes per second (we call this warm throughput) based on your previous traffic pattern, but your application may only be consuming 1,000 reads and 1,000 writes per second (consumed throughput). In this scenario, you are only charged for the 1,000 reads and 1,000 writes that you actually consume, even though at any time, your application could scale up to the war [source]
Instance pricing applies to both the primary and the replica instances in a cluster. Amazon DocumentDB utilizes a multi-Availability Zone (AZ) architecture for high availability. The cost of multi-AZ deployments is simply the cost of the primary instance plus the cost of each replica instance. [source] — Amazon DocumentDB bills primary and each replica
Database I/O: The amount of I/O used when reading and writing data to your cluster’s storage volume (pricing per million I/Os). [source]
I/Os are input/output operations performed by the Amazon DocumentDB engine when performing reads and writes against cluster’s storage volume. I/Os are billed per million I/O requests. API calls like find, insert, update, and delete, features like change streams and TTL indexes, and tools like mongodump and mongorestore all use I/Os by reading and/or writing from the storage volume. [source]
No. Amazon DocumentDB’s storage replication feature is included in the price. You are charged based on the size of the data in your cluster, and you are not charged separately for the replicated storage. [source] — Amazon DocumentDB storage replication included

Parameters and configuration

AWS's rule for DocumentDB: choose Standard if I/O is expected to be under 25% of total cluster cost, I/O-Optimized if over 25%. A cluster can switch to I/O-Optimized once every 30 days, and back to Standard at any time, with no downtime or reboot. [source] — Amazon DocumentDB storage choice

Examples and snippets

The on-demand pricing for I/O-Optimized db.r5.large instance is $0.3047 per hour, so your monthly cost for the two instances will be $444.86 ($0.3047 * 730 * 2). [source] — Amazon DocumentDB example 2: I/O-Optimized
For DocumentDB, due to high-IO rates, they opt for the I/O-Optimized plan. They evaluate the newest generation Graviton instance with 16 vCPUs and 128 GB of RAM, db.r5.4xlarge. The hourly cost is $2.44, making the monthly cost for one instance $1,756.8 ($2.44 x 720 hours). However, since it is a production database that needs to be highly available, they will need another instance in a different AZ, bringing the cost to $3,513.6 ($1756.8 x 2 instances). [source] — Vantage scenario; Graviton label for r5 looks wrong
Your total monthly bill for Amazon DocumentDB for this example would be $459.86 ($444.86 for instance + $15.00 for storage + $0.00 for I/O). In this example, with I/O-Optimized storage configuration, you saved 24.5% of your costs compared to standard configuration. [source] — Compares examples with different I/O volumes
The on-demand pricing for db.r5.large instance is $0.277 per hour, so your monthly cost for the two instances will be $404.42 ($0.277 * 730 * 2). [source] — Amazon DocumentDB example 1: Standard
Assume that your Amazon DocumentDB cluster is running a two-instance db.r5.large cluster (on-demand) in US East (N. Virginia) with a standard storage configuration. You store 50 GB of data with 50 GB backup and perform 200 million I/Os per month. [source] — Amazon DocumentDB example 1: Standard, 200M I/Os
Assume that your Amazon DocumentDB cluster is running a two-instance db.r5.large cluster (on-demand) in US East (N. Virginia) with I/O-Optimized storage configuration for your clusters. You store 50 GB of data with 50 GB backup and perform 1,000 million I/Os per month. [source] — Amazon DocumentDB example 2: I/O-Optimized, 1,000M I/Os
For MongoDB Atlas, they look at the M60 cluster which has 16 vCPU, 64 GB of RAM, and 320 GB of storage included. The hourly cost is $3.95. For a month with 720 hours (24 hours x 30 days) the total monthly cost is $2,844 ($3.95 x 720 hours). [source] — Vantage scenario: Atlas M60
A local mid-sized e-commerce company running a high-traffic online store is evaluating MongoDB Atlas and Amazon DocumentDB for their database needs. Their workload includes frequent product catalog updates, customer order processing, and real-time inventory tracking. Given their traffic patterns and need for high availability, they plan to deploy a 16 vCPU instance with 300 GB of storage in a multi-AZ setup. [source] — Vantage scenario
Your I/O cost will be $0.20 per million request increments, so for your 200 million I/Os per month, your cost will be $40.00 ($0.20 * 200). [source] — Amazon DocumentDB example 1: Standard
Over a month, we issue database requests consuming a total of 500,000 Request Units. The cost will be $0.125 (500,000 * $0.25/million). [source] — Azure Cosmos DB serverless example
Throughput bill for container in West US 10K RU/sec * 24 * 30 $0.008 per 100 RU/sec per hour $576 [source] — Azure Cosmos DB multi-region example, single write region
Throughput bill for 3 additional regions - East US, North Europe, and East Asia 3 * 10K RU/sec * 24 * 30 $0.008 per 100 RU/sec per hour $1,728 [source] — Azure Cosmos DB multi-region example, single write region
Throughput bill for container in West US (all regions are writable) 10K RU/sec * 24 * 30 $0.016 per 100 RU/sec per hour $1,152 [source] — Azure Cosmos DB multi-region example, multi-region writes
Throughput bill for 3 additional regions - East US, North Europe, and East Asia (all regions are writable) 3 * 10K RU/sec * 24 * 30 $0.016 per 100 RU/sec per hour $3,456 [source] — Azure Cosmos DB multi-region example, multi-region writes
Finally, the 300 GB of storage for Document DB is charged at $.30 per GB-month, corresponding to $90 ($.30 x 300 GB) in monthly storage costs and bringing the total to $3,603.6. [source] — Vantage scenario
Your storage cost will be $0.10 per GB/month, so for 50 GB storage, your (30-day) monthly cost will be $5.00 ($0.10 * 50). There will be no additional charge for your backup storage because it does not exceed your data storage for the region. [source] — Amazon DocumentDB example 1: Standard
Your storage cost will be $0.30 per GB/month, so for 50 GB storage, your (30-day) monthly cost will be $15.00 ($0.30 * 50). There will be no additional charge for your backup storage because it does not exceed your data storage for the region. [source] — Amazon DocumentDB example 2: I/O-Optimized
Storage bill for container in West US 250 GB $0.25/GB $62.50 [source] — Azure Cosmos DB multi-region example, single write region
Storage bill for 3 additional regions - East US, North Europe, and East Asia 3 * 250 GB $0.25/GB $187.50 [source] — Azure Cosmos DB multi-region example, single write region
Your total monthly bill for Amazon DocumentDB for this example would be $449.42 ($404.42 for instance + $5.00 for storage + $40.00 for I/O). [source] — Amazon DocumentDB example 1: Standard

Measurements and reference values

Azure Cosmos DB (NoSQL) bills provisioned throughput hourly at the maximum RU/s in that hour in 100 RU/s units, at about $0.008 per 100 RU/s per hour for a single write region and $0.016 with multi-region writes (US example rates). Storage is about $0.25/GB-month. Serverless mode bills $0.25 per million RUs consumed. [source] — Azure Cosmos DB provisioned and serverless rates
DocumentDB also has Serverless billed per DCU-hour ($0.0822 Standard, $0.0905 I/O-Optimized, 0.5 DCU minimum) and Elastic Clusters billed at $0.132 per vCPU-hour plus $0.30/GB-month storage, per the AWS pricing page. [source] — Amazon DocumentDB Serverless and Elastic Clusters
AWS cut DynamoDB on-demand throughput prices by 50% and global tables prices by up to 67% (replicated writes: 67% on-demand, 33% provisioned), effective 2024-11-01 in all Regions. AWS now recommends on-demand as the default mode and says most provisioned-capacity workloads get a lower price on on-demand. [source] — DynamoDB price cut, effective 2024-11-01
Cosmos DB reserved capacity gives up to 30% off for 1- or 3-year terms (Microsoft's worked example uses 20%), usable as a monetary credit across regions and APIs. [source] — Azure Cosmos DB reserved capacity
Amazon DocumentDB I/O-Optimized (I/O included configuration): Amazon DocumentDB I/O-Optimized is designed to prioritize price predictability or I/O intensive applications. If you expect your I/O costs to exceed 25% of your database cluster spend, this configuration offers enhanced price performance. With Amazon DocumentDB I/O-Optimized storage configuration, you are billed across three dimensions: on-demand instances, database storage, and backup storage. You are not charged for database I/O operations, ensuring predictable costs each billing cycle. [source]
Global tables pricing has been reduced by up to 67% – Building globally distributed applications is now significantly more affordable. DynamoDB has reduced pricing for multi-Region replicated writes to match the pricing of single-Region writes, simplifying cost modeling and choosing the best architecture for your applications. For on-demand tables, this price change lowers replicated write pricing by 67%, and for tables using provisioned capacity, replicated write pricing has been reduced by 33%. [source] — DynamoDB global tables
Starting from Amazon DocumentDB 5.0, instance-based clusters support two storage configuration types: + **Amazon DocumentDB standard storage**: Designed for customers with low to moderate I/O consumption. If you expect your I/O costs to be less than 25% of your total Amazon DocumentDB cluster, this choice might be ideal for you. With the Amazon DocumentDB standard storage configuration, you’re billed on a pay-per-request I/O basis in addition to instance and storage charges. This means your billing might vary from one cycle to another based on usage. The configuration is tailored to accommodat [source]
Our continued engineering investments on how efficiently we can operate DynamoDB allow us to identify and pass on cost savings to you. Effective November 1, 2024, DynamoDB has reduced prices for on-demand throughput by 50% and global tables by up to 67%, making it more cost-effective than ever to build, scale, and optimize applications. [source] — DynamoDB price cut
On-demand throughput pricing has been reduced by 50% – DynamoDB on-demand mode is now even more attractive, offering you a fully managed, serverless database experience that automatically scales in response to application traffic with no capacity planning required. On-demand mode’s capabilities like pay-per-request pricing, scale-to-zero, and no up-front costs help you save time and money while simplifying operations and improving performance at any scale. On-demand is a game changer for modern, serverless applications because it instantly accommodates workload requirements as they ramp up or [source] — DynamoDB on-demand
We have made DynamoDB even more cost-effective by reducing prices for on-demand throughput by 50% and global tables by up to 67%. Whether you are developing a new application, expanding to a global audience, or optimizing your cloud costs, the new DynamoDB pricing offers increased flexibility and substantial savings. [source] — DynamoDB price cut
Amazon DocumentDB Standard (pay-per-use I/O configuration): Amazon DocumentDB standard configuration is designed for low to moderate I/O consumption, and provides adaptability to suit your dynamic I/O needs. If you expect your I/O costs to be less than 25% of your database cluster spend, this configuration might be ideal for you. You are billed across all four dimensions: on-demand instances, database I/O, database storage, and backup storage. [source]
Let's consider that you've bought reserved capacity instead. You can buy reserved capacity for 100-K RU/sec at the price of $56,064 for one year (at 20% discount), or $6.40 per hour. See reserved capacity pricing on the Pricing page). [source]
Amazon DocumentDB T3 and T4g medium instances run in unlimited mode, which means that you will be charged if your average CPU utilization over a rolling 24-hour period exceeds the baseline of the instance. CPU Credits are charged at $0.09 per vCPU-Hour. The CPU Credit pricing is the same for all T3 and T4 instance sizes across all regions. [source]
Pricing for extended support for Amazon DocumentDB version 3.6 is based on instance family and vCPU-hours (which vary by AWS Region) and the number of years since standard support ended, adding an 80% premium for Year 1 and Year 2 and a 160% premium for Year 3. [source]
With Azure Cosmos DB free tier, you'll get the first 1000 RU/s and 25 GB of storage in your account for free, applied at the account level. Any RU/s and storage beyond 1000 RU/s and 25 GB will be billed at the regular pricing rates per the pricing page. On the bill, you will not see a charge or line item for the free 1000 RU/s and 25 GB, only the RU/s and storage beyond what is covered by free tier. To learn more, see how to create a free tier account article. [source]
Data replication: Amazon DocumentDB storage is highly durable and available, making your data durable across three Availability Zones (AZs) within a Region, you pay only for a single copy, with pricing as low as $0.10 GB/month (prices may vary across AWS regions). [source]
Storage capacity is billed in units of the maximum hourly amount of data stored, in GB, over a monthly period. For example, if you utilized 100 GB of storage for half of the month and 50 GB for the second half of the month, you would be billed for an equivalent of 75 GB of storage during that month. [source]
Cost of throughput (with reserved capacity) $70,080 discounted at 20% = $56,064 [source]
Storage Rate$0.10 per GB-month$0.30 per GB-month [source] — Amazon DocumentDB storage rates, us-east-1
There is no additional charge for backup storage of up to 100% of your total Amazon DocumentDB Elastic Clusters storage for the Region. [source]
Azure DocumentDB uses a straightforward compute and storage pricing model where both dimensions scale independently, with no throughput unit calculations or licensing fees on top. Backups are included at no extra charge for 35 days. [source] — Azure DocumentDB pricing model

Problems, failure modes and limitations

AWS's page example for two db.r5.large instances in US East shows $449.42/month on Standard (200M I/Os) and $459.86/month on I/O-Optimized (1,000M I/Os). The advertised 24.5% saving only holds when both are compared at the same I/O volume, so the totals on the page are not directly comparable. [source] — Amazon DocumentDB page example; I/O volumes differ

Comparisons and alternatives

Vantage (2025-02-11) modeled a 16 vCPU, 300 GB multi-AZ workload: Atlas M60 at $3.95/h (about $2,844/month, 320 GB included) vs DocumentDB db.r5.4xlarge I/O-Optimized at $2.44/h per instance, two instances plus storage totalling $3,603.60/month, making DocumentDB 26.7% dearer. Main driver: Atlas includes its replica set in the tier price while DocumentDB bills each instance. [source] — Vantage model; Amazon DocumentDB vs Atlas
A Microsoft Cosmos DB dev blog (2026-02-26) claims Azure DocumentDB (the Mongo-compatible vCore service) is 39-70% cheaper per hour than Atlas on AWS, e.g. M30-equivalent $0.296/h vs $0.570/h, M60 $2.369 vs $4.180. The article does not state currency, region or commitment terms, and the table is vendor-authored. [source] — Microsoft-authored; Azure DocumentDB vs Atlas on AWS
The two public Atlas-vs-competitor comparisons reach opposite conclusions (Microsoft: competitor cheaper; Vantage: Atlas cheaper) because they compare different competitor services (Azure DocumentDB vs AWS DocumentDB), different HA assumptions and different years. Neither is a neutral basis for a TCO claim; model the customer's own workload instead. [source]
While MongoDB Atlas includes multi-AZ replication at no extra cost, DocumentDB charges for each additional secondary instance, making high-availability deployments more expensive. [source] — Vantage claim
Amazon DocumentDB Serverless automatically scales database capacity up or down to meet your application workload needs. It offers cost savings for variable workloads, compared to provisioning instances for peak capacity. Capacity is measured in DocumentDB Capacity Units (DCUs) billed per second. 1 DCU has approximately 2 GiB of memory with a corresponding CPU and networking, similar to what is used in Amazon DocumentDB provisioned instances. [source]
When evaluating the two, one of the biggest deciding factors is cost. In a production scenario, we found DocumentDB to be 26.7% more expensive than MongoDB Atlas. [source] — Vantage claim
MongoDB Atlas has a catalog of useful features, multi-cloud deployment options, and predictable pricing. DocumentDB provides ease of use and seamless integration with AWS services. When comparing the pricing between the two for production workloads where high-availability is a requirement, we found MongoDB Atlas to be the cost-effective choice. [source] — Vantage conclusion
Usage-based pricing – Unlike provisioned capacity mode, where you pay for a fixed amount of throughput regardless of usage, with on-demand mode’s simple, pay-per-request pricing model, you don’t have to worry about idle capacity because you only pay for the capacity you actually use. You are billed per read or write request, so your costs directly reflect your actual usage. [source]
Amazon DocumentDB is 26.7% more expensive than MongoDB Atlas in this scenario. [source] — Vantage result
As a fully managed cloud-native database service, Azure Cosmos DB simplifies billing by charging only for your database operations and consumed storage. There are no additional license fees, hardware, utility costs, or facility costs compared to on-premises or IaaS-hosted alternatives. When you consider the multi region capabilities of Azure Cosmos DB, the database service provides a substantial reduction in costs compared to existing on-premises or IaaS solutions. [source]
Amazon recommends multi-AZ deployments for high availability, however, unlike MongoDB Atlas, you are charged for it. Costs are per primary instance plus each secondary instance. [source] — Vantage article
MongoDB Atlas is the cost-effective choice for production workloads where high-availability is a requirement. [source] — Vantage claim
Customers can achieve meaningful cost savings compared to managed MongoDB offerings on AWS, driven by predictable vCore pricing and independent compute and storage scaling [source] — Microsoft-authored claim

Changes and history

These pricing changes are already in effect, in all Regions, starting November 1, 2024, and will be automatically reflected in your monthly AWS bill. We’re excited about what these changes mean for customers and the value you can realize from DynamoDB. For more details, see Pricing for Amazon DynamoDB. [source] — DynamoDB price cut
Editor’s note (May 2026): This article was published in the Azure Cosmos DB Dev Blog before the launch of the dedicated Azure DocumentDB Dev Blog. Azure DocumentDB now has its own home for product announcements, engineering insights, migration guidance, technical deep dives, and developer tutorials. If you’re building with Azure DocumentDB or the open-source DocumentDB project, we recommend bookmarking the Azure DocumentDB Dev Blog for the latest news and technical content. [source] — Microsoft editor note, May 2026

Facts and statements

Amazon DocumentDB Serverless instantly scales databases to support hundreds of thousands of transactions per second and supports all features of Amazon DocumentDB, including Multi-AZ deployments and read replicas. You can choose to configure all the instances in your database cluster to use the DocumentDB Standard or I/O- Optimized storage configuration based on the price performance and price predictability needs of your application. To learn more, you can review the Amazon DocumentDB Serverless pricing examples. [source]
On-Demand vCPUs let you pay per second, with no long-term commitments or upfront fees. This frees you from the cost and complexity of planning and purchasing database capacity in advance, and from having to guess the correct capacity. Pricing is per vCPU hour consumed, from the time your Amazon DocumentDB Elastic Clusters is launched until it is deleted. vCPUs provide RAM (memory), I/O operations, and the required CPU cores for your workload. Partial vCPU hours are billed in one-second increments, with a 10-minute charge following a billable status change such as creating, modifying, or deleti [source]
You can add/remove Azure regions anywhere in the world to your Azure Cosmos DB database account at any time. The throughput that you have configured for various Azure Cosmos DB databases and containers will be reserved in each of the Azure regions associated with your Azure Cosmos DB database account. If the sum of provisioned throughput (RU/sec) configured across all the databases and containers within your Azure Cosmos DB database account (provisioned per hour) is T and the number of Azure regions associated with your database account is N, then the total provisioned throughput for a given h [source]
Provisioned Throughput: You are billed hourly for the maximum provisioned throughput for a given hour, in increments of 100 RU/s. [source]
Serverless: You are billed hourly for the total amount of Request Units consumed by your database operations. [source]
You pay for replicated write I/O operations between the primary region and each secondary region. The number of replicated write I/Os to each secondary region equals the number of in-region write I/Os performed by the primary region. These replicated write I/O charges apply to both Amazon DocumentDB Standard and Amazon DocumentDB I/O-Optimized cluster storage configurations. In addition to replicated write I/Os, you pay for instances, storage, cross-region data transfer, backup storage, and I/O usage in both the primary and secondary regions based on your chosen cluster storage configuration. [source]
Storage configuration: Choose between standard and I/O-Optimized storage configurations for your database cluster, enabling you to maximize your price performance. [source]
Like MongoDB Atlas, Amazon DocumentDB charges based on instance type while instances are active. It offers 29 instance options, including burstable types for lower-cost workloads and high-performance options for demanding applications. The price depends on region and the chosen pricing model—Standard or I/O-Optimized. I/O-Optimized costs more for compute and storage, but does not charge for I/O requests. It is more cost-effective for cases where I/O costs exceed 25% of database cluster spend. [source] — Vantage article
On-demand instances let you pay per second, with no long-term commitments or upfront fees. This frees you from the cost and complexity of planning and purchasing database capacity in advance and having to guess the correct capacity. [source]
Extended Support for Amazon DocumentDB version 3.6 begins on March 31, 2026, and usage billing will start on July 1, 2026. For example, if you are running a db.r4.large instance (2 vCPUs) on Amazon DocumentDB version 3.6 in the US East (Ohio) region, the normal charge is $0.277 per hour. Extended Support adds the following charges: July 1, 2026 to March 31, 2028: $0.222 per hour ($0.111 per vCPU-hour × 2 vCPUs) and April 1, 2028 to March 31, 2029: $0.444 per hour ($0.222 per vCPU-hour × 2 vCPUs). [source]
If you know typical request rates for your current database workload, read about estimating request units using Azure Cosmos DB capacity planner [source]
Charges also apply for data transfer, backup storage, and optional features such as Elastic clusters. [source] — Vantage article
Other AWS-wide costs such as data transfer between applications and Amazon DocumentDB across Availability Zones (AZs) may also apply. [source]
With Amazon DocumentDB Serverless, the minimum database capacity you can set is 0.5 DCUs. The database scales instantly from 0.5 DCUs to 7 DCUs, runs for 30 minutes, and scales down to the minimum capacity of 0.5 DCUs. The database goes from 7 DCUs to 0.5 DCUs in approximately 3 minutes. Therefore, the following is the compute cost for running this workload. [source]
Storage consumed by your Amazon DocumentDB cluster is billed per GB-month. Billable storage includes your data, indexes, and change stream data. For more information, see Amazon DocumentDB Storage. [source]
You pay only for the storage that your Amazon DocumentDB Elastic Clusters consumes and don’t need to provision storage in advance. Storage consumed by your Amazon DocumentDB Elastic Clusters is billed per GB-month. Billable storage includes your data, indexes, and change stream data. For more information, see Amazon DocumentDB Storage. [source]
Note: Extended Support charges are in addition to standard DocumentDB pricing for instances, storage, I/O, and backup storage. [source]
You can switch your existing database clusters once every 30 days to Amazon DocumentDB I/O-optimized storage. You can switch back to Amazon DocumentDB standard storage at any time. The next date to modify the storage configuration to I/O-optimized can be tracked with the `describe-db-clusters` command using the AWS CLI or through the AWS Management Console in the cluster's configuration page. [source]
You can create a new database cluster including the Amazon DocumentDB I/O-optimized configuration or convert your existing database clusters with a few clicks in the [AWS Management Console](https://console.aws.amazon.com/docdb/), a single parameter change in the [AWS Command Line Interface (AWS CLI)](https://aws.amazon.com/cli/), or through [AWS SDKs](https://aws.amazon.com/developer/tools/). There is no downtime or reboot of instances required during or after modifying the storage configuration. [source]
Database Savings Plans: A flexible pricing model that offers savings in exchange for a commitment to a specific amount of usage (measured in $/hour) over a 1-year term. You can learn more about eligible usage on the Database Savings Plans pricing page. [source] — Amazon DocumentDB
Pricing is per instance-hour consumed, from the time an instance is launched until it is stopped or deleted. Partial instance hours are billed in one-second increments, with a 10-minute minimum charge following a billable status change such as creating, modifying, or deleting an instance. [source]
Multiplying 10 units per hour by the cost of $0.008 (per 100 RU/sec per hour) = $0.08 per hour. [source]
If you increase provisioned throughput at 9:30 AM from 400 RU/sec to 1,000 RU/sec and then lower provisioned throughput at 10:45 AM back to 400 RU/sec, you will be charged for two hours of 1,000 RU/sec. [source]
On-demand instances: The amount of compute instances for a cluster (pricing per second with a 10-minute minimum). [source]
Backup storage: The amount of backup storage used in excess of your cluster’s database storage usage (pricing per GB/month). [source]
Low-cost backup: You get backup storage equivalent to 100 percent of your cluster’s data storage for free each month (e.g., if your cluster has 10 TB of stored data, you get 10 TB of backup for free). Additional backup storage beyond the free allotment is priced as low as $0.02 per GB/month (prices may vary across AWS regions). [source]
vCPU: The amount of compute measured in vCPUs for a cluster (priced per minute with a 10-minute minimum). Database storage: The amount of data stored in your cluster’s storage volume (pricing per GB-month). Backup storage: The amount of backup storage used in excess of your cluster’s database storage usage (pricing per GB-month). [source]
There is no additional charge for backup storage of up to 100 percent of your total Amazon DocumentDB cluster storage for a region. There is also no additional charge for backup storage if your backup retention period is one day and you do not have any manual snapshots beyond the backup retention period. [source]
Backup storage beyond this amount and manual snapshots retained outside of the backup retention period, whether your cluster is active or has been deleted, are billed at storage rates in the table below. [source]
To ensure high availability MongoDB Atlas automatically deploys clusters across two or three AWS Availability Zones (AZs) (depending on how many AWS AZs there are for the selected region). [source] — Vantage article on Atlas
Per-second billing: Amazon DocumentDB provides per second billing for instances, with a 10-minute minimum billing period. [source]
The pricing below is based on data transferred “in” and “out” of Amazon DocumentDB. [source]
Database operations: The way you get charged for your database operations depends on the type of Azure Cosmos DB account you are using. [source]
Multiplying the $0.08 per hour by the number of hours in the month equals $0.08 * 24 hours * 30 days = $57.60 for the month. [source]

Disagreements

A Microsoft Cosmos DB dev blog (2026-02-26) claims Azure DocumentDB (the Mongo-compatible vCore service) is 39-70% cheaper per hour than Atlas on AWS, e.g. M30-equivalent $0.296/h vs $0.570/h, M60 $2.369 vs $4.180. The article does not state currency, region or commitment terms, and the table is vendor-authored. [source] — Microsoft-authored; Azure DocumentDB vs Atlas on AWS
When evaluating the two, one of the biggest deciding factors is cost. In a production scenario, we found DocumentDB to be 26.7% more expensive than MongoDB Atlas. [source] — Vantage claim
Amazon DocumentDB is 26.7% more expensive than MongoDB Atlas in this scenario. [source] — Vantage result
Customers can achieve meaningful cost savings compared to managed MongoDB offerings on AWS, driven by predictable vCore pricing and independent compute and storage scaling [source] — Microsoft-authored claim

Related concepts

pricing — is a related of Cross-Vendor Document Database Pricing Comparison (DocumentDB Cosmos DB DynamoDB)
DocumentDB — is a part of Cross-Vendor Document Database Pricing Comparison (DocumentDB Cosmos DB DynamoDB)
on-demand — is a measure of Cross-Vendor Document Database Pricing Comparison (DocumentDB Cosmos DB DynamoDB)
instance-hour — is a measure of Cross-Vendor Document Database Pricing Comparison (DocumentDB Cosmos DB DynamoDB)
storage cost — is a measure of Cross-Vendor Document Database Pricing Comparison (DocumentDB Cosmos DB DynamoDB)
Cosmos DB — is a part of Cross-Vendor Document Database Pricing Comparison (DocumentDB Cosmos DB DynamoDB)
DynamoDB — is a part of Cross-Vendor Document Database Pricing Comparison (DocumentDB Cosmos DB DynamoDB)
I/O-Optimized — is a part of Cross-Vendor Document Database Pricing Comparison (DocumentDB Cosmos DB DynamoDB)
provisioned throughput — is a measure of Cross-Vendor Document Database Pricing Comparison (DocumentDB Cosmos DB DynamoDB)
Atlas — is a instance of Cross-Vendor Document Database Pricing Comparison (DocumentDB Cosmos DB DynamoDB)
multi-AZ — is a part of Cross-Vendor Document Database Pricing Comparison (DocumentDB Cosmos DB DynamoDB)
Request Unit — is a measure of Cross-Vendor Document Database Pricing Comparison (DocumentDB Cosmos DB DynamoDB)
savings percentage — is a measure of Cross-Vendor Document Database Pricing Comparison (DocumentDB Cosmos DB DynamoDB)
reserved capacity — is a measure of Cross-Vendor Document Database Pricing Comparison (DocumentDB Cosmos DB DynamoDB)
price cut — is a problem of Cross-Vendor Document Database Pricing Comparison (DocumentDB Cosmos DB DynamoDB)
each replicated region — is a part of Cross-Vendor Document Database Pricing Comparison (DocumentDB Cosmos DB DynamoDB)
Vantage — is a instance of Cross-Vendor Document Database Pricing Comparison (DocumentDB Cosmos DB DynamoDB)
included replica set — is a contrast of Cross-Vendor Document Database Pricing Comparison (DocumentDB Cosmos DB DynamoDB)
name clash — is a problem of Cross-Vendor Document Database Pricing Comparison (DocumentDB Cosmos DB DynamoDB)

Children

DynamoDB on-demand vs provisioned break-even (frontier, viability 2.60 / 5, below research gate)
Cosmos DB RU estimation for MongoDB workloads (frontier, viability 2.45 / 5, below research gate)
Atlas included-replica pricing vs per-instance HA billing (frontier, viability 2.65 / 5, below research gate)

Frontier under this node: Atlas included-replica pricing vs per-instance HA billing (2.65), Cosmos DB RU estimation for MongoDB workloads (2.45), DynamoDB on-demand vs provisioned break-even (2.60)

← the whole tree · 3D view· how to read this page