I am trying to run 10 short sequences (20 bases) against a custom database of 17 genomes, where I want the results to be only 20 bp long and allow for up to 3 mismatches. I am running:

blastn -task blastn-short -query example_seq.fasta -db Example_db -out C:\blastn_test.csv -outfmt "10 sseqid qseqid sstart send sstrand evalue sseq length btop score" -word_size 11 -qcov_hsp_perc 75

This is the closest I can get to achieving an output that I want. This result shows alignments but not at the length I want them to be. I have written simple code in R using Biostrings that easily matches these sequences, but I wanted to use BLAST to make things easier and it is seeming far from it at the minute. It seems the local blast through command line cannot specify how many mismatches to account for like you can with R or Biopython.

Am I missing something completely?

Any help would be appreciated, apologies if I haven't posted correctly, this is my first post.