Hi Eric,

you can use UCSC genome browser.

For batch processing you can read the FAQ`s or use Galaxy.

I hope this helps.

If you're into programming (and R (and bioconductor)), let's assume you're in human:

## Setting the environment:
library(BSgenome.Hsapiens.UCSC.hg19)
library(GenomicRanges)

## Let's say these are some of the genomic coordinates
## of your non-coding RNAs:
coords <- GRanges(c('chr1', 'chr2', 'chr3'),
                  IRanges(c(1e7, 2e7, 3e7), width=100),
                  c('+', '-', '+'))
## which looks like:
coords
GRanges with 3 ranges and 0 elementMetadata values:
      seqnames               ranges strand
         <Rle>            <IRanges>  <Rle>
  [1]     chr1 [10000000, 10000099]      +
  [2]     chr2 [20000000, 20000099]      -
  [3]     chr3 [30000000, 30000099]      +

## There is an `Hsapiens` object loaded into your workspace
## which got there by loading the
## BSgenome.Hsapiens.UCSC.hg19 library, you can combine a
## GRanges object with `Hsapiens` to easily fetch the seqences

seqs <- getSeq(Hsapiens, coords)

## Which look like:
seqs
  A DNAStringSet instance of length 3
    width seq
[1]   100 AACCCCGTCTCTACAATAAATTAAAATATTAGCT...CTTGGCGGGCTGAGGTGGGAGAATCATCCAAGC
[2]   100 GCAGGTTTGGAATCTGTTTCATGAAAACTCCCAG...TCATCCATTTGCATAGCATCAAATGGAAATGTC
[3]   100 ATGTTTGATTGATGCCCTCGAACTTGGCTAGTTG...AAACTCAATTGAATTATATTGCCTTAGCTTATA

While convenient, the getSeq function might require more ram than iterating over the coords of your loci one chromosome at a time, but I'll leave those implementation details as an assignment for the reader ;-)