On this page
Atlas data transfer costs depend on the cloud service provider hosting your database deployment. Atlas tabulates data transfer costs daily.
Multi-region clusters might have higher data transfer costs depending on the number and location of additional regions, as well as the number of clusters deployed to each region.
Exceptions to Data Transfer Cost
Atlas doesn't charge for data transfer for incoming or outgoing
M5 cluster tiers.
Serverless instances incur costs for data transfer to and from the virtual machine responsible for backing up and restoring data.
MongoDB hosts and operates the free Atlas Live Migration Service to help users migrate existing MongoDB databases to MongoDB Atlas. MongoDB doesn't charge for any incoming data transfers to an Atlas database deployment. Learn more about migrating to Atlas.
The vast majority of Atlas customers spend less than 10% of their budget on data transfer. If you are spending significantly more, some of these optimizations may reduce your data transfer costs:
Check all applications and processes that access your data for inefficiencies. Ensure that queries do not:
Re-read data that already exists on the client.
Re-write existing data to your database deployment.
Ensure that queries originate from the same cloud region and provider as your database deployment whenever possible.
When cross-region queries are necessary:
Use the Aggregation Framework to preprocess your data before you transfer it. For example, you can project document fields using the
$projectaggregation stage to reduce the size of a document before you transfer it.
Ensure that your client driver uses wire protocol compression to communicate with MongoDB. Atlas always compresses intra-cluster communication. To learn how to configure your driver, refer to your Driver's Documentation.
Queries from on-premises environments into Atlas, across cloud providers, or between continents on the same cloud provider incur the greatest data transfer costs.