Split PDF by multiple pages using PDFTK?

You can use sejda-console, it's open source under AGPLv3 and can be downloaded from the project GitHub page.

You can use the splitbyevery command which

Splits a given PDF document every 'n' pages creating documents of 'n' pages each.

In you case the command line will be something like:

sejda-console splitbyevery -n 2 -f /tmp/input_file.pdf -o /out_dir

This PowerShell script will

1. use pdftk to get the number of pages
2. loop in steps building a range string
3. use the range to extract the pages into a new pdf with appended range to the base name (and store in the same folder).

Change the first two vars to fit your environment.

## Q:\Test\2017\05\06\Split-Pdf.ps1
$pdfPath = 'Q:\Test\2017\05\06\'
$pdfFile = Join-Path $pdfPath "test.pdf"
$SetsOfPages = 3
$Match = 'NumberOfPages: (\d+)'
$NumberOfPages = [regex]::match((pdftk $pdfFile dump_data),$Match).Groups[1].Value
"{0,2} pages in {1}" -f $NumberOfPages, $pdfFile

for ($Page=1;$Page -le $NumberOfPages;$Page+=$SetsOfPages){
  $File = Get-Item $pdfFile
  $Range = "{0}-{1}" -f $page,[math]::min($Page+$SetsOfPages-1,$NumberOfPages)
  $OutFile = Join-Path $pdfPath ($File.BaseName+"_$Range.pdf")
  "processing: {0}" -f $OutFile
  pdftk $pdfFile cat $Range output $OutFile
}

Edited to work with variable sets of pages and to properly handle the overhang.
Edited again: found a much easier way do shorten the last set of pages.

Sample output

> .\Split-Pdf.ps1
10 pages in Q:\Test\2017\05\06\test.pdf
processing: Q:\Test\2017\05\06\test_1-3.pdf
processing: Q:\Test\2017\05\06\test_4-6.pdf
processing: Q:\Test\2017\05\06\test_7-9.pdf
processing: Q:\Test\2017\05\06\test_10-10.pdf