Amazon S3
Hacker ApprovedThe most recommended of 10 object storage
A highly reliable, scalable object store with a simple, widely copied API that is cheaper than building your own, though its API has accumulated quirks and costs can surprise.
A highly reliable, scalable object store with a simple, widely copied API that is cheaper than building your own, though its API has accumulated quirks and costs can surprise.
254 people gave an opinion
- recommend it
- 170
- criticize it
- 84
- approval share
- 67%
Recommended in 12 of 13 months
- November 2025: recommended
- December 2025: recommended
- January 2026: recommended
- February 2026: recommended
- March 2026: recommended
- April 2026: recommended
- May 2026: recommended
- June 2026: recommended
- July 2026: recommended
- August 2026: recommended
- September 2026: recommended
- October 2026: recommended
- November 2026: not recommended (month in progress)
Owned for 14 years
Median of the 3 owners who said how long they have had it.
Each square is one year.
Why it carries the badge
- 170 people recommend itNeeds 10 or more, met.
- 2 recommenders for every criticNeeds 2 to 1 or better, met.
- Recommended in 12 of 13 monthsNeeds 4 or more, met.
The same three rules apply to every product, and the badge can’t be bought. How the badge works
What people say about it
The reasons people give, with the number of comments that give each one. Open a reason to read what they said.
Why people recommend it
Each mark is one comment
Extremely reliable and durable
“it's a extremely reliable, infinitely-scalable, strongly consistent (in some implementation), cheap, and extremely simple to use key-value store”
Comment from 1 Oct 2026 in “Cloudflare K2: serverless event streams” “S3 is and was an amazing resource for normal people. Extremely reliable cloud storage for backup, for distribution, for anything”
Comment from 8 Sep 2026 in “Among European Companies That Use a CDN, Nearly 9 in 10 Use Cloudflare” “Their metrics are, quite honestly, absolutely impossible for most engineering teams and companies to ever achieve”
Comment from 21 Aug 2026 in “Git at Any Scale” - Comment from 28 Sep 2026 in “S3 Is the Future, S3 Is the Past”
- Comment from 14 Nov 2025 in “650GB of Data (Delta Lake on S3). Polars vs. DuckDB vs. Daft vs. Spark”
- Comment from 16 Jun 2026 in “What job interviews taught me about Kubernetes”
Cheaper than building it yourself
“The nice thing about S3 is even if you screw up your usage patterns, you can pay/engineer your way out guaranteed.”
Comment from 15 Nov 2025 in “AWS deprecates two dozen services (most of which you've never heard of)” “It's cheaper to pay S3 to deal with the problem than to have a full team dealing with bit rot from RAID issues.”
Comment from 12 Oct 2026 in “Postgres isn't slow, your storage is” “S3 is still magnitudes cheaper, faster, and more redundant than anything you could put together yourself”
Comment from 27 Sep 2026 in “S3 Is the Future, S3 Is the Past” - Comment from 14 Nov 2025 in “650GB of Data (Delta Lake on S3). Polars vs. DuckDB vs. Daft vs. Spark”
- Comment from 18 Dec 2025 in “Your job is to deliver code you have proven to work”
Simple API, de facto standard
“it's a extremely reliable, infinitely-scalable, strongly consistent (in some implementation), cheap, and extremely simple to use key-value store”
Comment from 1 Oct 2026 in “Cloudflare K2: serverless event streams” “Either you use something like the S3 interface, or you waste your time implementing a clone of what S3 already supports.”
Comment from 20 Dec 2025 in “TP-Link Tapo C200: Hardcoded Keys, Buffer Overflows and Privacy” “The range of things you can do with blob storage and a (very simple) auth model are surprisingly broad.”
Comment from 1 Oct 2026 in “Cloudflare K2: serverless event streams” - Comment from 28 Sep 2026 in “S3 Is the Future, S3 Is the Past”
- Comment from 14 Apr 2026 in “I Just Want Simple S3”
Conditional writes enable coordination
“The quiet improvements to S3 over the last 3 years like conditional puts unlocked a lot of use cases not previously feasible”
Comment from 13 Aug 2026 in “Celld: Self-hosted, distributed Durable Objects” “the coordination layer/control-plane is all S3 CAS of dumb json files, which is a fantastic common-mode infra requirement”
Comment from 5 Aug 2026 in “Celld: Self-hosted, distributed Durable Objects” “S3 is strongly consistent, if you need it anyways you can just use s3 keys to deconflict and store the workflow state.”
Comment from 29 May 2026 in “SQLite is all you need for durable workflows” - Comment from 24 Feb 2026 in “A distributed queue in a single JSON file on object storage”
What people warn about
Each mark is one comment
API shows accumulated cruft
“Originally the first S stood for "Simple". With all the additions they've had to bolt on, S2 would be far more appropriate a name.”
Comment from 3 Dec 2025 in “MinIO is now in maintenance-mode” “Another weird thing is that `foo/` is a valid object name in s3, so you can have both `foo/` and `foo` be different files.”
Comment from 19 Apr 2026 in “It is incorrect to "normalize" // in HTTP URL paths” “decision to make bucket (and not bucket + account id surrogate) a sole key for access was one of most annoying mistakes”
Comment from 13 Mar 2026 in “Bucketsquatting Is (Finally) Dead” - Comment from 3 Dec 2025 in “MinIO is now in maintenance-mode”
- Comment from 3 Dec 2025 in “MinIO is now in maintenance-mode”
Pricing and surprise charges
“For reference transit cost has dropped over the years and is now about at $0.000247/GB or free if there is a peering agreement”
Comment from 27 Sep 2026 in “S3 Is the Future, S3 Is the Past” “AWS have found the ideal captive market: computer engineers who don't know how computers work or how much they should cost.”
Comment from 1 Oct 2026 in “Cloudflare K2: serverless event streams” “trying to fix the monthly bleed with a Lifecycle Policy nearly had a seven figure charge on top of it all”
Comment from 20 Apr 2026 on Hacker News - Comment from 28 Sep 2026 in “S3 Is the Future, S3 Is the Past”
Key prefix quirks
“Another weird thing is that `foo/` is a valid object name in s3, so you can have both `foo/` and `foo` be different files.”
Comment from 19 Apr 2026 in “It is incorrect to "normalize" // in HTTP URL paths” “there were also some implementation specifics like throughput limits based on key prefixes (that one vanished circa 2016)”
Comment from 7 Apr 2026 in “S3 Files and the changing face of S3” “Even teams that use prefixes for performance don't have an S3-native way to ask what a prefix represents”
Comment from 4 May 2026 in “Buckets and objects are not enough”
Other object storage people recommend
The four most recommended other object storage, with the same counts for each, so you can compare like for like. The tick on each bar marks the 2 to 1 line the badge needs.
| Product | Recommend | Criticize | Approval | Months | Owned | Badge |
|---|---|---|---|---|---|---|
| Amazon S3This page | 170 | 84 | 67% | 12 | 14 yrs3 owners | Approved |
| Cloudflare R2 | 85 | 14 | 86% | 12 | — | Approved |
| Backblaze B2 | 50 | 14 | 78% | 12 | — | Approved |
| AWS S3 | 17 | 7 | 71% | 11 | — | Approved |
| Amazon S3 Glacier Deep Archive | 16 | 11 | 59% | 7 | — | — |
What people said
40 of the 302 comments that mention it, newest first. Each links to the original comment.
“It's cheaper to pay S3 to deal with the problem than to have a full team dealing with bit rot from RAID issues.”
“AWS have found the ideal captive market: computer engineers who don't know how computers work or how much they should cost.”
“it's a extremely reliable, infinitely-scalable, strongly consistent (in some implementation), cheap, and extremely simple to use key-value store”
“The range of things you can do with blob storage and a (very simple) auth model are surprisingly broad.”
“it is advertised as eleven nines durability, however. That is better than anything you can build yourself”
“Why did you claim that the high price is to cover the costs if 99.9% of it isn't covering the costs?”
“I finally stopped using (S)FTP when i realized that all the ftp clients now have native support for S3”
“For reference transit cost has dropped over the years and is now about at $0.000247/GB or free if there is a peering agreement”
“S3 is still magnitudes cheaper, faster, and more redundant than anything you could put together yourself”
“I use S3 (well GCS mostly) for data that I wouldn’t be upset if it’s lost. And I pay the zonal discount for it.”
“S3 is and was an amazing resource for normal people. Extremely reliable cloud storage for backup, for distribution, for anything”
“a cloud backup (think s3 or azure blob) with credentials that can't be harvested automatically by a malware”
“Their metrics are, quite honestly, absolutely impossible for most engineering teams and companies to ever achieve”
“The quiet improvements to S3 over the last 3 years like conditional puts unlocked a lot of use cases not previously feasible”
“the coordination layer/control-plane is all S3 CAS of dumb json files, which is a fantastic common-mode infra requirement”
“the cloud provider managed services for e.g. blob stores or databases are so much more reliable”
“That's how far s3 has come - it's replacing products that were the pinnacle of scalability back in the day.”
“S3 is strongly consistent, if you need it anyways you can just use s3 keys to deconflict and store the workflow state.”
“S3 is pretty competitive if you want similarly-performing storage with consistent millisecond-level latency”
“Even teams that use prefixes for performance don't have an S3-native way to ask what a prefix represents”
“trying to fix the monthly bleed with a Lifecycle Policy nearly had a seven figure charge on top of it all”
“Another weird thing is that `foo/` is a valid object name in s3, so you can have both `foo/` and `foo` be different files.”
“It gives a nice limited surface area that doesn't allow you to do things that can get too complex”
“there were also some implementation specifics like throughput limits based on key prefixes (that one vanished circa 2016)”
“It can be unprofessional to design a system that takes an external variable like S3 for granted, especially if it's not needed.”
“decision to make bucket (and not bucket + account id surrogate) a sole key for access was one of most annoying mistakes”
“S3 CAS conditional writes changed the game for building coordination primitives on object storage”
“S3-compatible object storage as the substrate is a nice escape from the operational complexity of running Redis or SQS”
“a myriad of intermediate steps are introduced to write data to dynamo/s3/kinesis + sqs/and coordination”
“S3 also gives you multipart parallel uploads. Each part gets stored and then when you're done the concatenation is performed close to the storage layer.”
“the site didn't need to be updated once built so serving a load of static files via S3 makes ongoing maintenance very low”
“There are no alternatives that support most of the AWS S3 functionality such as replication, event notifications.”
“Once in S3, triggers start a series of lambdas to decrypt, decompress, analysis. Works well, easy to reason about.”
“Either you use something like the S3 interface, or you waste your time implementing a clone of what S3 already supports.”
“the way our software works is 95% cheaper to run using S3 as the storage rather than DynamoDB”
“I suspect they learned a lot over the years and the API shows the scars. In their defense, they did go first.”
“Originally the first S stood for "Simple". With all the additions they've had to bolt on, S2 would be far more appropriate a name.”
“Everything uses poorly documented, sometimes inconsistent HTTP headers that read like afterthoughts/tech debt.”
“The nice thing about S3 is even if you screw up your usage patterns, you can pay/engineer your way out guaranteed.”
“S3 is an amazingly engineered product, operates at truly impressive scale, is quite reasonably priced”
Counts come from 1,795,640 Hacker News comments, November 2025 to November 2026. How we count. Report a mistake.