Delimiter isn’t used for writes, only list operations. S3 simply looks at the common string prefixes in your object names and uses that to internally shard obj

Delimiter isn’t used for writes, only list operations. S3 simply looks at the common string prefixes in your object names and uses that to internally shard objects, so you can achieve a multiple of those request limits. aaa122348 aaa484585 bbb484858 bbb474827 Would have same performance as: aaa/122348 aaa/484585 bbb/484858 bbb/474827