AWS adds S3 Files

Amazon Web Services has turned Amazon Simple Storage Service buckets into mountable file systems, letting applications use file reads, writes, and locks instead of object-storage calls. (aws.amazon.com) Amazon announced Amazon S3 Files on April 7, 2026 and said the feature is generally available in 34 Amazon Web Services Regions. The company said customers can attach an S3 file system to compute resources and see changes reflected back to the underlying bucket. (aws.amazon.com) Amazon Simple Storage Service was built as object storage, which stores data as discrete objects in buckets rather than in folders and files that operating systems expect. Amazon S3 Files adds file-system behavior on top, so applications can open directories, update files, and use file locking without moving the data out of S3. (docs.aws.amazon.com 1) (docs.aws.amazon.com 2) Amazon said the service is aimed at workloads that have long sat awkwardly between cheap object storage and more interactive file storage, including analytics jobs, artificial intelligence agents, and shared data pipelines. In its launch post, Amazon said multiple compute resources can attach to the same data set with about 1 millisecond latency and without duplicating data. (aws.amazon.com) Under the hood, Amazon says S3 Files is built using Amazon Elastic File System, the company’s managed network file system. The documentation says it keeps a low-latency storage layer for active file data and metadata while leaving the source data in S3. (docs.aws.amazon.com 1) (docs.aws.amazon.com 2) The system does not pull an entire bucket into file storage at once. Amazon says S3 Files imports directory metadata when a directory is first listed or a file is opened, and by default also imports file data smaller than 128 kilobytes. Larger data is brought in as workloads touch it. (docs.aws.amazon.com) Amazon also tied the launch to its container and compute stack. The company’s documentation says users can mount S3 Files on Amazon Elastic Compute Cloud instances, and Amazon Elastic Container Service can expose S3 Files volumes directly inside task definitions. (docs.aws.amazon.com 1) (docs.aws.amazon.com 2) Amazon is pitching that as a way to keep one copy of data in S3 while older file-based software and newer artificial intelligence tools both use it through familiar file paths. The company’s launch materials describe S3 Files as an attempt to remove the tradeoff between S3’s scale and cost profile and the file semantics many applications still require. (aws.amazon.com)