# Gen1 vs Gen2 IOPS billing > Source-anchored research for Gen1 vs Gen2 IOPS billing. Run status: COMPLETED. Parent: [Storage Growth and IOPS Forecasting for Cost](https://llms-explorer.com/tree/storage-growth-and-iops-forecasting-for-cost/) · 7 facets · 17 facts · page: https://llms-explorer.com/tree/gen1-vs-gen2-iops-billing/ ## Definitions - Negation finding: switching to Gen2 is not evidence of a lower bill. MongoDB says generation, IOPS type, and IOPS rate affect cost and recommends estimating each change in the Atlas cluster builder. It also recommends changing generation and IOPS type separately and checking performance between changes. (high confidence) (verified as of 2026-10-24) — [source](https://www.mongodb.com/docs/atlas/reference/faq/storage.md) ## Structure and components - At the underlying provider layer, AWS gp3 charges provisioned IOPS above its 3,000-IOPS baseline separately from storage capacity. Google Hyperdisk Balanced generally includes 3,000 IOPS of baseline performance and bills provisioned performance above baseline; volumes in Advanced performance Hyperdisk Storage Pools do not receive an included baseline. These are provider pricing rules, not Atlas unit prices or Atlas's storage-size-dependent IOPS baseline. (medium confidence) (verified as of 2026-10-24) — [source](https://aws.amazon.com/ebs/pricing/) ## How-to and procedures - Use Atlas configuration quotes and billing records to price Atlas IOPS; do not substitute EBS or Hyperdisk list prices. The cloud documents specify their own included baselines and billable provisioned performance, while Atlas defines its own storage-dependent Gen2 baseline and Extended Standard IOPS charge. (medium confidence) (verified as of 2026-10-24) — [source](https://www.mongodb.com/docs/atlas/billing/cluster-configuration-costs/) ## Measurements and reference values - Atlas documents a Gen2 baseline of 3,000 IOPS below 1,000 GB and a 3:1 IOPS-to-GB ratio above 1,000 GB. Extended Standard IOPS ceilings are 80,000 on AWS and 160,000 on Google Cloud; actual supported settings depend on the cluster configuration. (high confidence) (verified as of 2026-10-24) — [source](https://www.mongodb.com/docs/atlas/customize-storage/) - Atlas documents a 3,000-IOPS Gen2 baseline below 1,000 GB, and a 3:1 IOPS-to-GB baseline above 1,000 GB; its examples include 1,000 GB with 3,000 IOPS, 2,000 GB with 6,000 IOPS and 4,000 GB with 12,000 IOPS. These are Atlas baselines, not direct cloud-volume pricing rules. (low confidence) (verified as of 2026-10-24) — [source](https://www.mongodb.com/docs/atlas/customize-storage/) - AWS EBS gp3 includes 3,000 provisioned IOPS and 125 MB/s baseline throughput; additional IOPS and throughput are independently provisioned and billed. This describes direct EBS pricing and does not supply an Atlas IOPS unit rate. (low confidence) (verified as of 2026-10-16) — [source](https://aws.amazon.com/ebs/pricing/) ## Problems, failure modes and limitations - This comparison applies to eligible M30+ Atlas Core clusters on AWS or Google Cloud. M10 and M20 are generation-agnostic. Atlas Infinite uses Gen2 but does not support Extended Standard IOPS, so generation alone cannot identify the applicable IOPS cost model. (high confidence) (verified as of 2026-10-24) — [source](https://www.mongodb.com/docs/atlas/cluster-config/change-cluster-generation.md) ## Comparisons and alternatives - Changing Gen1 to Gen2 preserves storage capacity and IOPS rate; on AWS it also preserves IOPS type. A retained rate can exceed the Gen2 storage-included baseline, so review the IOPS configuration and resulting price after migration rather than assuming automatic savings. (high confidence) (verified as of 2026-10-24) — [source](https://www.mongodb.com/docs/atlas/cluster-config/change-cluster-generation.md) - Outside commentary from LASSIC describes AWS Gen2 storage and IOPS as independently scalable and separately billable; ISSOH describes the same independent controls. Both attribute technical details to MongoDB documentation, so they are discovery leads rather than independent measurements of Atlas billing. (medium confidence) (verified as of 2026-10-02) — [source](https://lassic.co.jp/media/column/mongodb-atlas-cost-optimization-outsourcing/) - On AWS, Gen1 Provisioned IOPS and Gen2 High Performance IOPS name the same performance category. They differ from Standard/Extended Standard IOPS; Gen2 does not automatically convert a retained Provisioned IOPS configuration to Standard IOPS. (high confidence) (verified as of 2026-10-24) — [source](https://www.mongodb.com/docs/atlas/customize-storage/) ## Facts and statements - For Atlas Core Gen2 Dedicated clusters, storage charges include baseline IOPS; Extended Standard IOPS provisioned above that baseline incur a separate charge. Gen1 standard IOPS scale with storage, while Gen2 permits independent changes to standard IOPS and storage capacity. (high confidence) (verified as of 2026-10-24) — [source](https://www.mongodb.com/docs/atlas/billing/cluster-configuration-costs/) - AWS Gen1 Provision IOPS becomes Gen2 High Performance IOPS, while Gen1 Standard remains Gen2 Standard. Extended Standard IOPS is a different option from High Performance IOPS. Standard uses general-purpose SSDs; High Performance uses provisioned-IOPS SSDs with different latency and consistency characteristics. (high confidence) (verified as of 2026-10-24) — [source](https://www.mongodb.com/docs/atlas/reference/faq/storage.md) - Atlas documentation says M30+ Dedicated clusters on AWS or Google Cloud can use Gen1 or Gen2, with different IOPS, storage, and cost for the same cluster class. For Gen2, the disk-size baseline IOPS are included with storage and Extended Standard IOPS above baseline are charged separately; a Gen1-to-Gen2 switch maintains storage capacity and IOPS rate, and AWS maintains the IOPS type. (medium confidence) (verified as of 2026-10-24) — [source](https://www.mongodb.com/docs/atlas/billing/cluster-configuration-costs/) - For Atlas Core, Gen1 Standard IOPS and storage scale together; Gen2 allows independent scaling of Standard IOPS and storage. Gen2 Dedicated clusters are an M30+ option on AWS and Google Cloud, subject to supported configurations. (high confidence) (verified as of 2026-10-24) — [source](https://www.mongodb.com/docs/atlas/manage-clusters/) - Gen2 storage charges include the baseline minimum IOPS for the disk size. Atlas charges separately for Extended Standard IOPS provisioned above that baseline; independent IOPS scaling is not a promise that all IOPS are free. (high confidence) (verified as of 2026-10-24) — [source](https://www.mongodb.com/docs/atlas/billing/cluster-configuration-costs/) - MongoDB scopes IOPS options by provider, generation, and cluster class: AWS Gen1 uses Provisioned IOPS and Gen2 uses High Performance IOPS; Gen2 AWS or Google Cloud clusters can use Extended Standard IOPS. The storage guide says Gen1 M40+ clusters can customize IOPS only in General and Low CPU classes, not Local NVMe SSD, and the generation-change guide says AWS preserves IOPS type across migration. (medium confidence) (verified as of 2026-10-24) — [source](https://www.mongodb.com/docs/atlas/billing/cluster-configuration-costs/) - Google says Hyperdisk Balanced volumes include a baseline of 3,000 IOPS and 140 MiB/s throughput; provisioned performance above those baselines is billable. That baseline applies to Standard performance, Standard capacity, and Advanced capacity Hyperdisk Storage Pools, but not Advanced performance pools. This describes Google infrastructure pricing, not an Atlas billing schedule. (low confidence) (verified as of 2026-10-22) — [source](https://docs.cloud.google.com/compute/docs/disks/hd-types/hyperdisk-balanced)