Older versions of the engine appended a numeric suffix such as _0 to the generated archive name, which made downloads look like reports_0.zip. That behaviour is gone: the archive name now comes from the job itself.
Control it with:
zipFileName— the name used when the archive is downloaded (for examplereports.zip).zipTo— the destination key in the bucket where the archive is stored.
If you still see a suffix, check that you are not passing a name that already exists at that key — a fresh destination keeps the name you asked for.