Skip to main content
Cloud & AI Hub
Browse
Glossary AI Directory Playgrounds Models Prompts Explainers Strategy Matrix Benchmark Decoder

Glacier Deep Archive

AWS's cheapest storage tier — roughly $1 per terabyte per month — for data you must keep but almost never read, with retrieval measured in hours.

Last reviewed: July 25, 2026

What is Glacier Deep Archive?

S3 Glacier Deep Archive is AWS’s lowest-cost storage class: around **0.00099perGB−month∗∗—roughly0.00099 per GB-month** — roughly 1 per terabyte per month, ~23× cheaper than S3 Standard. The trade for that price is retrieval latency: getting data back takes up to 12 hours (standard) and incurs per-request and per-GB retrieval fees. It’s engineered for a specific shape of data — must be retained, will almost never be read — and it’s the correct home for exactly that.

The workloads it’s built for

Compliance and legal retention (financial records held 7+ years, healthcare archives, regulatory data you’re required to keep and hope never to touch), long-term backups you’d only restore in a disaster, and raw data archives kept “just in case.” The mental model: Deep Archive is tape, priced and behaving like tape — cheap to store, slow and metered to retrieve. If there’s any chance you’ll need the data within minutes, it’s the wrong tier.

Where the costs hide

The sticker price is real but incomplete. Retrieval isn’t free: pulling a large archive back incurs both per-request and per-GB retrieval charges, so a “cheap” archive you restore frequently isn’t cheap. Minimum storage duration: objects deleted before 180 days are billed as if stored the full 180 — churny data in Deep Archive pays a penalty. Retrieval time is a planning constraint, not a detail: a disaster-recovery plan that assumes instant restore from Deep Archive is broken; the 12-hour window must fit your RTO. Lifecycle policies automate the migration — data lands in Standard, ages to Infrequent Access, then to Deep Archive on a schedule.

What people get wrong

  • Archiving frequently-accessed data and paying retrieval fees that dwarf the storage savings.
  • DR plans assuming fast restore — Deep Archive’s hours-long retrieval breaks aggressive recovery objectives.
  • Ignoring the 180-day minimum and getting billed for storage you deleted after a week.

Primary source: Amazon S3 Glacier storage classes

Deep Archive’s Retrieval Time Tradeoff

Glacier Deep Archive offers the lowest per-gigabyte storage cost of any AWS storage tier, but this comes with a meaningful tradeoff in retrieval time: standard retrievals typically take within 12 hours to complete, a stark contrast to S3 Standard’s effectively instant access. This makes Deep Archive appropriate specifically for data that genuinely needs to be retained (often for regulatory or compliance reasons) but is expected to be accessed rarely, if ever — long-term backup archives, regulatory records with multi-year retention requirements, or historical data kept purely for potential future audit purposes — rather than for any data that might need to be accessed on short notice, where a less extreme cold-storage tier would be a more appropriate fit despite its somewhat higher cost.

Advertisement (In-Content)

Historical figures and technical concepts for informational purposes only. Not technical, professional, legal, or financial advice. Sources: Official Documentation.