EFS Pricing: Throughput Modes, Storage Classes, Lifecycle

Amazon EFS occupies an awkward billing position. It is more expensive per GB than any block-storage or object-storage alternative on AWS, by a wide margin — $0.30/GB-month for Standard storage is more than 13× S3 Standard, 4× gp3 EBS storage, and even ahead of the io2 Block Express premium tier. The premium pays for shared POSIX semantics that no other AWS storage primitive offers natively. When that premium is justified, EFS is the right answer. When it isn’t, the bill becomes one of the largest unjustified line items in a typical AWS account.

Amazon EFS pricing bills storage class, throughput mode, and provisioned throughput independently. In us-east-1 (June 2026), Standard storage is $0.30/GB-month while Standard-IA is $0.025/GB-month. Most surprises come from keeping cold data on Standard tier and paying burst throughput on Provisioned mode when Elastic would suffice.

Storage class	us-east-1	Access pattern
Standard	$0.30/GB-month	Hot shared files
Standard-IA	$0.025/GB-month	under 3% access frequency
Archive	$0.008/GB-month	Rarely accessed
One Zone	~50% off Standard	Non-critical dev data

This post is the bill story. For the architecture decision of whether to use EFS, Lambda S3 Files, or stay on S3 entirely, our Lambda S3 Files vs EFS guide covers the trade-offs. For S3-side pricing math, the S3 pricing calculator is a good reference point.

The Seven EFS Billing Dimensions

EFS pricing breakdown — us-east-1, June 2026

Prices in us-east-1

Storage class × throughput mode is the matrix. Pick wrong at provision time and the bill is locked in until you migrate.

Dimension	Unit price	Example workload	Monthly cost
Standard storage Multi-AZ; the most expensive tier	$0.30 / GB-month	1 TB hot dataset	$300.00
Standard-IA 92% cheaper at rest; pay per access	$0.025 / GB-month + $0.01/GB read	1 TB rarely-accessed archive	$25.00 + reads
Archive 99% cheaper; 90-day minimum storage	$0.0036 / GB-month + $0.03/GB read	1 TB compliance hold	$3.60 + reads
One Zone Standard 47% cheaper than multi-AZ; single-AZ durability	$0.16 / GB-month	1 TB regenerable training data	$160.00
One Zone-IA 95% cheaper than Standard; single-AZ + cold	$0.0133 / GB-month	1 TB cold regenerable data	$13.30
Elastic Throughput Default for new file systems; bursty-friendly	$0.03/GB read, $0.06/GB write	500 GB read + 100 GB write / month	$21.00
Provisioned Throughput Fixed cost; for consistently-high throughput	$6.00 / MBps-month	100 MBps provisioned	$600.00

Standard storage

$300.00

Multi-AZ; the most expensive tier

Unit price: $0.30 / GB-month
Example workload: 1 TB hot dataset

Standard-IA

$25.00 + reads

92% cheaper at rest; pay per access

Unit price: $0.025 / GB-month + $0.01/GB read
Example workload: 1 TB rarely-accessed archive

One Zone Standard

$160.00

47% cheaper than multi-AZ; single-AZ durability

Unit price: $0.16 / GB-month
Example workload: 1 TB regenerable training data

One Zone-IA

$13.30

95% cheaper than Standard; single-AZ + cold

Unit price: $0.0133 / GB-month
Example workload: 1 TB cold regenerable data

Elastic Throughput

$21.00

Default for new file systems; bursty-friendly

Unit price: $0.03/GB read, $0.06/GB write
Example workload: 500 GB read + 100 GB write / month

Provisioned Throughput

$600.00

Fixed cost; for consistently-high throughput

Unit price: $6.00 / MBps-month
Example workload: 100 MBps provisioned

Lifecycle transitions: $0.01/GB Standard → IA, $0.03/GB Standard → Archive. Recovered within the first month of cold storage for most files.

If you only audit one dimension first, enable Intelligent-Tiering — it is the highest-impact one-click optimization on EFS.

The 13× Premium: Why EFS Storage Costs So Much

EFS Standard costs more per GB than any other AWS storage primitive. The premium is not arbitrary — it pays for:

Multi-AZ durability with automatic replication across availability zones in a region, with no application-side configuration.
POSIX semantics including file locking (flock, fcntl), atomic renames, hard links, and standard Unix permissions.
Concurrent access from any compute primitive that mounts NFSv4.1 — Lambda, EC2, ECS, EKS, Fargate, on-premises via Direct Connect.
Elastic capacity that scales without provisioning; no resize operation, no provisioned-IOPS decision, no “filling up” condition.

The premium is genuinely worth it for workloads that need these properties: content-management systems with concurrent author writes, ML training pipelines with shared dataset access from many GPU workers, legacy applications that expect a mounted filesystem, scientific computing with HPC-style shared scratch.

It is not worth it for workloads that don’t use these properties. Read-mostly training data, model artifacts, reference datasets, cache layers, and stateless batch processing all routinely sit on EFS for historical reasons rather than because they need POSIX semantics.

Standard-IA and Archive: The 92–99% Storage Saving Hiding in Plain Sight

EFS Intelligent-Tiering, enabled at the file-system level, automatically moves files between Standard and Standard-IA (and Archive) based on access patterns. Without it, every file sits on Standard at $0.30/GB-month forever, even files no one has touched in three years.

The numbers on a 5 TB file system with a typical hot/cold distribution:

5 TB EFS file system: with and without Intelligent-Tiering

Prices in us-east-1

Typical access pattern: 20% of files account for 80% of access; the rest is cold.

Dimension	Unit price	Example workload	Monthly cost
No tiering — all on Standard Status quo without lifecycle policy	$0.30 / GB-month	5 TB × $0.30	$1,500.00
Standard + IA (Intelligent-Tiering) 73% saving; small read-fee tail	~80% on IA, 20% on Standard	1 TB Standard + 4 TB IA	$400.00
Standard + IA + Archive 76% saving; appropriate for old data	20% Standard, 40% IA, 40% Archive	1 TB Standard + 2 TB IA + 2 TB Archive	$357.20
IA read fees (5% of IA accessed) Read fees are small in well-tuned tiering	$0.01 / GB read	200 GB read/month from IA	+$2.00

No tiering — all on Standard

$1,500.00

Status quo without lifecycle policy

Unit price: $0.30 / GB-month
Example workload: 5 TB × $0.30

Standard + IA (Intelligent-Tiering)

$400.00

73% saving; small read-fee tail

Unit price: ~80% on IA, 20% on Standard
Example workload: 1 TB Standard + 4 TB IA

Standard + IA + Archive

$357.20

76% saving; appropriate for old data

Unit price: 20% Standard, 40% IA, 40% Archive
Example workload: 1 TB Standard + 2 TB IA + 2 TB Archive

IA read fees (5% of IA accessed)

+$2.00

Read fees are small in well-tuned tiering

Unit price: $0.01 / GB read
Example workload: 200 GB read/month from IA

Enabling Intelligent-Tiering is a single console toggle. The transition happens automatically; the saving applies to new accesses going forward.

The single-click enabling of Intelligent-Tiering is one of the highest-impact cost optimizations available on AWS. Most file systems have not enabled it.

Elastic Throughput: The 2024 Default That Changed the Pricing Math

Until 2024, EFS throughput was either Bursting (credit-based; throttled under sustained load) or Provisioned ($6/MBps-month; fixed cost regardless of actual usage). Elastic Throughput, introduced in 2023 and the default for new file systems since 2024, replaces both with a per-GB-transferred model: $0.03 per GB read, $0.06 per GB write, scaling instantly to workload demand.

For most workloads, Elastic is cheaper than Provisioned. A workload averaging 50 GB/day read and 20 GB/day write costs about $1.5/day in read fees ($45/mo) plus $1.2/day in write fees ($36/mo) — $81/month total — versus provisioning 100 MBps at $600/month even when the actual peak is much lower than the provisioned rate.

Where Provisioned still wins: workloads with sustained high throughput (above ~25 MBps continuously) where the per-GB Elastic charges accumulate faster than the fixed Provisioned monthly fee.

One Zone Storage: Half the Price for Half the Durability

One Zone storage classes (One Zone Standard at $0.16/GB-month, One Zone-IA at $0.0133/GB-month) store data in a single availability zone rather than replicated across multiple AZs. The price reflects the reduced durability: an AZ failure means data unavailability until the AZ recovers, and a true regional AZ destruction means data loss.

For regenerable data — model training datasets that can be re-downloaded from a primary source, scratch space, cache layers, ML model artifacts that re-publish from a CI/CD pipeline — One Zone is correct and saves 47–95% on storage. For primary production data, multi-AZ is the right insurance and the premium is worth it.

The waste pattern: defaulting to multi-AZ for all EFS file systems out of habit, including ones whose data is trivially regenerable.

Lambda S3 Files: The 13× Cheaper Alternative for the Right Workload

We covered the architectural decision in our Lambda S3 Files vs EFS cost and limits post. The pricing summary: S3 Standard is $0.023/GB-month against EFS Standard’s $0.30/GB-month — a 13× difference. For workloads where Lambda S3 Files’ six documented limitations (no flock(), eventual consistency on writes, throughput caps, etc.) are acceptable, the storage saving alone justifies the migration on any file system over a few hundred GB.

The decision tree from that post:

Does the workload need POSIX file locking? → Stay on EFS.
Does the workload need strict write consistency? → Stay on EFS.
Does the workload need sustained high write throughput? → Stay on EFS (or evaluate FSx for Lustre for HPC patterns).
Otherwise → Lambda S3 Files (for Lambda-mounted access) or direct S3 SDK (for API-driven access).

For the GPU training dataset case specifically, S3 with a Mountpoint for S3 client often outperforms EFS at a fraction of the cost.

Common EFS Bill Surprises

When EFS Is and Is Not the Right Storage

EFS for workloads that genuinely need shared POSIX semantics; S3 (with or without Mountpoint/S3 Files) for read-mostly; gp3 for single-writer block storage; FSx for HPC.

Use when

Content management or collaborative authoring systems with concurrent writes requiring file locking
ML training pipelines where many workers need POSIX file-system access to shared dataset
Legacy applications that expect a mounted filesystem and cannot be re-architected to use object storage
Multi-AZ shared scratch for batch processing where elastic capacity matters
Workloads moving across compute primitives (Lambda + EC2 + Fargate) sharing the same data

Avoid when

Read-mostly training datasets — S3 with Mountpoint for S3 or Lambda S3 Files is 13× cheaper
Single-writer database storage — EBS gp3 with snapshot-based DR is dramatically cheaper
Backup target — S3 Standard-IA or Glacier is the correct tier; EFS is not designed for cold storage
High-throughput sustained workloads where FSx for Lustre would be both faster and cheaper
Any workload where Standard-class is selected without enabling Intelligent-Tiering — pure waste

When EFS is the right answer, it is decisively right. When it is the wrong answer, the premium compounds month after month.

A 30-Day EFS Bill Cleanup Plan

Week 1 — Enable Intelligent-Tiering everywhere. Audit file systems via aws efs describe-file-systems --query "FileSystems[].LifecyclePolicies". Enable Intelligent-Tiering with a 30-day transition threshold on every file system over 100 GB. This typically recovers 60–80% of the storage line within 30 days as cold files transition.

Week 2 — Audit Provisioned Throughput. Identify file systems still on Provisioned Throughput via aws efs describe-file-systems --query "FileSystems[].ThroughputMode". For each, compare actual usage (CloudWatch BurstCreditBalance history and DataReadIOBytes / DataWriteIOBytes) against provisioned rate. Switch to Elastic Throughput where utilization is consistently below 80% of provisioned.

Week 3 — Identify migration candidates. For each remaining file system over 500 GB, classify the workload by access pattern. Single-writer database storage → migrate to EBS gp3. Read-mostly training data → migrate to S3 with Mountpoint or Lambda S3 Files. Sustained high-throughput HPC → consider FSx for Lustre. Document the decision per file system.

Week 4 — Backup retention. Audit AWS Backup plans backing EFS sources. Apply explicit retention policies. Move long-term retention to S3-based AWS Backup vaults rather than retaining EFS-side recovery points.

Cross-check the S3 alternative pricing in our S3 pricing calculator when scoping migration candidates.

What This Post Doesn’t Cover

EFS Replication (cross-region) — billed as a combination of source/destination storage and inter-region transfer; treat as a separate budget line.
EFS-on-FSx-for-Windows — different product entirely; pricing model is different (per-GB throughput-capacity).
Cross-account EFS access — uses AWS RAM and standard EFS pricing; no additional EFS-side charge.
In-depth FSx for Lustre comparison — covered in a separate HPC-focused post; relevant when sustained throughput requirements exceed what EFS provisioned mode can serve cost-effectively.

If You Only Do One Thing This Week

Enable Intelligent-Tiering on every EFS file system over 100 GB with the default 30-day transition threshold. The change is non-destructive, happens in the background, and typically recovers 60–80% of the storage line within 30 days. Pair with a quarterly review of FilesAccessedFromIA and MeteredIOBytes CloudWatch metrics to confirm the tiering ratio is working as expected. For file systems with stale data older than a year, also enable Archive-tier transition for a deeper saving on truly-cold files.

For the architectural decision of whether to stay on EFS at all, the Lambda S3 Files vs EFS guide lays out the decision tree.

Frequently asked questions

Is EFS really 13× more expensive than S3?

For storage alone, yes — EFS Standard at $0.30/GB-month vs S3 Standard at $0.023/GB-month. The premium reflects what EFS provides that S3 does not: POSIX semantics, file locking, concurrent multi-AZ access from any compute primitive that can mount NFS, and elastic capacity that scales without provisioning. The premium is worth it for workloads that genuinely need shared-filesystem semantics (training data with file-system access patterns, content management, legacy applications that expect a mount). It is hard to justify for read-mostly workloads that could use S3 with Lambda S3 Files, FSx, or just an object-store SDK.

How does Standard-IA actually save 92%?

Standard-IA stores files that have not been accessed for a configurable period (7, 14, 30, 60, 90 days; default 30) at $0.025/GB-month — 92% cheaper than Standard. The catch: every access of an IA file incurs a per-GB read fee ($0.01/GB) and a small latency penalty (first byte ~50–100ms vs sub-ms on Standard). For datasets with a clear hot/cold ratio — typical archive use cases sit at 80–95% cold — the all-in IA bill including occasional read fees still lands at 70–85% below pure Standard storage. EFS Intelligent-Tiering automates the placement; enable it on every file system that holds data more than 30 days old.

What is the EFS Archive tier and when does it pay off?

Archive is the coldest EFS tier at $0.0036/GB-month — 99% cheaper than Standard, 86% cheaper than Standard-IA. It is designed for files accessed less than once a quarter. Access fees are higher ($0.03/GB read), and there is a 90-day minimum storage duration before transition out. The right use case is compliance retention on file-system-shaped data — research datasets you must keep but realistically will never re-process, regulatory artifacts in file form. The wrong use case is operational data with even occasional access; the per-GB read fees compound quickly and the latency penalty hurts user-facing workflows.

Why did AWS change the default to Elastic Throughput?

Elastic Throughput, introduced in 2023 and the default for new file systems since 2024, eliminates the throughput-capacity provisioning decision. Old throughput modes (Bursting and Provisioned) required you to either accept a credit-based burst model that throttled under sustained load, or to provision throughput at $6/MBps-month — a high fixed cost. Elastic Throughput bills per GB transferred ($0.03 read, $0.06 write) regardless of read/write rate, scaling instantly to workload needs. The trade-off: workloads with consistent high throughput can be cheaper under provisioned mode; bursty or moderate workloads are cheaper under elastic. AWS defaulted to Elastic because most workloads are bursty.

When does One Zone storage make sense?

One Zone storage classes (One Zone Standard at $0.16/GB-month, One Zone-IA at $0.0133/GB-month) are 47–53% cheaper than the multi-AZ equivalents. The trade-off is durability — One Zone stores data in a single AZ, so an AZ failure means data unavailability until that AZ recovers. The right use case is replicable data — model training datasets that can be regenerated, cache layers, scratch space for batch processing — where the cost saving is material and reconstruction is cheap. The wrong use case is primary production data; the multi-AZ pricing is the right insurance.

Is Lambda S3 Files actually 13× cheaper than EFS for read workloads?

For read-mostly workloads with the right access pattern, yes — and we covered the architecture in our [Lambda S3 Files vs EFS post](/blog/aws-lambda-s3-files-vs-efs-cost-and-limits/). The S3-backed POSIX mount stores data at S3 Standard pricing ($0.023/GB-month) versus EFS Standard ($0.30/GB-month). The catch: Lambda S3 Files has six documented limitations (no file locking, eventual consistency on writes, throughput caps, etc.). For training data, model artifacts, reference datasets where the workload is read-mostly and stateless, the migration is straightforward; for workloads that depend on flock(), strict consistency, or high write throughput, EFS remains correct.

How does the IA lifecycle transition fee work?

EFS bills a one-time transition fee of $0.01/GB when a file moves from Standard to IA, and a one-time fee when it moves from Standard/IA to Archive. The transition fees are recovered within the first month of cold storage for most files (the storage saving is dramatically greater than the one-time transition). The exception: files that transition to IA and then get accessed (re-promoting them to Standard) generate both the transition fee and the read fee — and if the file then ages out again, another transition fee. For workloads with unpredictable access patterns, Intelligent-Tiering uses a built-in heuristic to avoid frequent transitions; explicit lifecycle policies should err on the side of longer transition thresholds (60–90 days rather than 7) for data with bursty access.

Amazon EFS Pricing: The Throughput Mode and Storage Class Decisions That Decide the Bill

The Seven EFS Billing Dimensions