Object storage
Also called: S3, bucket storage, blob storage
Object storage keeps data as flat keyed objects accessed over HTTP rather than as a filesystem, and is the usual destination for backups and large artifacts.
Object storage stores data as objects in a flat namespace, addressed by key over an HTTP API, rather than as files in a directory tree.
The trade-off against a filesystem: you give up partial writes, locking and rename, and get effectively unlimited capacity, durability across hardware failures, and access from anywhere with credentials. Objects are written and read whole.
Amazon S3 defined the API and almost everyone else implements it — Backblaze B2, Cloudflare R2, Hetzner Object Storage, and self-hosted MinIO or Ceph. Code written against S3 generally moves between them by changing an endpoint, which keeps the switching cost low.
Watch egress pricing: the per-gigabyte storage cost is often the smaller number, and the charge to read data back is what makes providers expensive at restore time.