Merge CSV files

The merge is refused. Every file after the first has its header compared cell by cell against the first file, and any difference stops the run with a message naming the file and saying that merging would misalign columns. Column order and spelling both count. Use Map CSV column names or Reorder / insert CSV columns to line them up first.

The order you dropped the files in, with the rows of each file kept in their original sequence. The header is written once at the top and every input header row is skipped. Nothing is sorted and nothing is deduplicated; run Remove duplicate CSV rows afterwards if you need that.

Yes. Turn hasHeader off and every row of every file is treated as data and concatenated straight through, with no header check. That is the right setting for files that are already headerless, and it also skips the mismatch protection, so make sure the columns line up.

merged.csv, always, with a result line saying how many files went in and how many data rows came out. If every input turned out to have zero data rows the tool refuses with "every input had 0 data rows" rather than writing a file with nothing but a header.

Yes. Each file is parsed with quote state tracked across the whole text, so a record with a newline inside a quoted cell counts as one row and stays one row. The merged output is re-quoted the same way, so cells containing commas, quotes or newlines come out valid.

There is no fixed limit. Every file is read into memory and parsed before the merged output is built, so the peak memory is roughly the sum of the inputs plus the result. Nothing is uploaded, so the practical ceiling is your device. Several hundred MB of CSV is more than a browser tab will enjoy.