Note
Access to this page requires authorization. You can try signing in or changing directories.
Access to this page requires authorization. You can try changing directories.
APPLIES TO:
MongoDB
Important
Are you looking to migrate an existing MongoDB application or use MongoDB Query Language (MQL) features? Consider Azure DocumentDB.
Are you looking for a database solution for high-scale scenarios with a 99.999% availability service level agreement (SLA), instant autoscale, and automatic failover across multiple regions? Consider Azure Cosmos DB for NoSQL.
Note
If you are planning a data migration to Azure Cosmos DB and all that you know is the number of vcores and servers in your existing sharded and replicated database cluster, please also read about estimating request units using vCores or vCPUs
Configuring your databases and collections with the right amount of provisioned throughput, or Request Units (RU/s), for your workload is essential to optimizing cost and performance. This article describes how to use the Azure Cosmos DB Cost Estimator to get an estimate of the required RU/s and cost of your workload when using the Azure Cosmos DB for MongoDB. If you're using API for NoSQL, see how to use the Azure Cosmos DB Cost Estimator with API for NoSQL.
The API for MongoDB operations map to the Cost Estimator's operation types as follows: find → Reads, insert → Writes, update → Updates, delete → Deletes, and aggregate/query → Queries.
To estimate the throughput and cost of your workload, go to the Azure Cosmos DB Cost Estimator. The Cost Estimator is in public preview, and you don't need to sign in. As you enter the inputs that describe your workload, the tool automatically updates the Projected monthly cost total. There's no separate calculate step.
Note
The Azure Cosmos DB Cost Estimator is in public preview. The projected cost is an estimate only and isn't a quote.
Start with a scenario or configure it yourself
When you open the Cost Estimator, choose between two paths:
- Guide me to my estimate (about 2 minutes): A guided wizard that walks you through a series of questions to build your estimate quickly.
- Let me configure for myself (about 5 minutes): Tune every input yourself in the Configure panel.
You can select Start over at any time to return to this choice.
Guide me to my estimate
The guided path is a nine-step wizard: Workload, Stage, Data, Traffic, Backup, Read/Write Pattern, Region, Review, and Estimate. Start by picturing your workload and asking whether one of the common scenario patterns fits. Choose from many patterns, such as:
- Chat and Sessions
- IoT Device Telemetry
- Orders and Transactions
- RAG and Embeddings
- Product Catalog and Inventory
If none of the patterns match, choose None of these fit to answer questions about your specific workload instead. The running estimate shows "TBD/mo" until enough information is provided.
Let me configure for myself
This path opens the Configure panel directly, where you can tune every input. The sections and fields are described in the next section.
Configure your estimate
In the Configure panel, work through each collapsible section and provide the inputs that match your workload. Each section includes More info (ⓘ) tooltips. To reset all inputs, select Start over in the upper-left corner.
The following table describes the key inputs. Operation rates use generic operation types: Reads, Writes, Updates, Deletes, and Queries. See the article that references this walkthrough for how your API's operations map to these types.
| Section / Input | Description |
|---|---|
| Traffic & Capacity – Capacity Mode | Choose Serverless, Autoscale, or Manual. Serverless bills per request, so your estimate depends on workload activity; add a data type in the Workload section to see the projected monthly cost. Autoscale provisions throughput that scales automatically with traffic and suits variable or unpredictable workloads. Manual provisions a fixed amount of throughput and suits sustained, predictable traffic. |
| Data – Storage | The total estimated data stored (GB) in a single region. Select a quick pick (1, 10, 100, or 1,000 GB) or enter a custom value in GB. |
| Data – Indexing Mode | Controls how properties are indexed, which affects the RU charge for writes. Automatic indexing indexes all properties in all items for flexible and efficient queries. Turning indexing off means none of the properties are indexed, which results in the lowest RU charge for writes, suitable when you only do key/value lookups and writes, and no queries. A custom indexing policy lets you tune which properties are indexed to balance write cost, storage, and query flexibility. |
| Workload – Per second / Per month toggle | Choose whether the operation rates you enter are expressed per second or per month. |
| Workload – Data type (documents) | Add one or more data types to represent the documents in your workload. Each data type has a Name, a document size (quick picks 1, 5, 20, or 100 KB, or a custom value), and a # of Docs count. Use Add data type for workloads that contain multiple document shapes, and Import JSON or View JSON to work from sample documents. |
| Workload – Operation rates | For each data type, enter the expected rate of Reads, Writes, Updates, Deletes, and Queries. The tool includes built-in RU hints for each operation type. |
| Workload – Query complexity | Select the complexity that best matches your queries: simple (about 3 RU), filter (about 10 RU), or complex (about 30 RU). |
| Backup & Consistency – Backup Mode | Select the backup mode for your account. |
| Backup & Consistency – Consistency Level | Azure Cosmos DB supports five consistency levels so you can balance the tradeoffs between consistency, availability, and latency. By default, Azure Cosmos DB uses session consistency, which guarantees the ability to read your own writes in a session. Choosing strong or bounded staleness requires roughly double the RU/s for reads compared to session, consistent prefix, and eventual consistency. Strong consistency with multi-region writes isn't supported. |
| Region – Region and Region Count | Azure Cosmos DB is available in all Azure regions. Select a region and the number of regions required for your workload. For more information, see Distribute your data globally with Azure Cosmos DB. |
| Region – Multi-Region Writes | If you enable multi-region writes, your application can read and write to any Azure region. If you disable multi-region writes, your application writes to a single region. Enable multi-region writes for active-active workloads that require low-latency writes in different regions, for example, an IoT workload that writes at high volume across regions. Multi-region writes guarantee 99.999% read and write availability and require more throughput than single write regions. For more information, see Optimize multi-region cost in Azure Cosmos DB. |
| Region – Availability Zones | Enable availability zones to improve resiliency within a region. |
| Pricing – Discounts | Apply pricing options that affect your estimate: Free Tier, Reserved Capacity (None, 1-Year, or 3-Year), and an Enterprise Agreement discount percentage. |
| Advanced – Vector Operations | Enable if your app uses AI or semantic search (for example, "find similar products"). Turning it on adds storage for vector embeddings plus RU/s for vector write and query operations. When enabled, you set Vector Dimensions (500, 768, 1536, or 3072), Index Type (Flat, QuantizedFlat, or DiskANN), Vector Writes/sec, Vector Queries/sec, and Partitions scanned per query. |
| Advanced – Dedicated Gateway | Enable to model an integrated cache. You choose a Gateway SKU (for example, D4s, D8s, or D16s) and the number of Gateway Instances. |
| Advanced – Change Feed Consumers | Enable to account for the cost of processes that read the change feed, and set the number of consumers. |
Review and export your estimate
As you enter inputs, the Projected monthly cost updates live. Select Summary at the top of the tool to see the running total. The Summary shows the cost "Based on" the selected capacity mode, with a breakdown of the throughput and storage that drive the cost.
Select Compare to see how cost breaks down across throughput, storage, and backup. The Compare view includes:
- An itemized cost breakdown table showing each item, its configuration, formula, monthly cost, and percent of total.
- A throughput by document type table showing each document type with its size, operation rates, RU/s, and cost per month.
- A cost projection chart that models 12 months of growth with adjustable storage-growth and compute-growth rates.
Use Compare to view different configurations side by side. For example, compare serverless versus provisioned throughput, or single-region versus multi-region.
When your estimate is ready, use Export to save it, or select Create in Azure Portal to start provisioning an account based on your inputs.
The prices shown in the Azure Cosmos DB Cost Estimator are estimates based on the public pricing rates for throughput and storage. All prices are shown in US dollars. To see all rates by region, see the Azure Cosmos DB pricing page.
Next steps
- 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
- Learn more about Azure Cosmos DB's pricing model.
- Create a new Azure Cosmos DB account, database, and container.
- Learn how to optimize provisioned throughput cost.
- Learn how to optimize cost with reserved capacity.