The EMBOSS program cusp takes one or more nucleotide sequences as input and outputs codon usage data, looking like this (first few lines):

#CdsCount: 1

#Coding GC 67.79%
#1st letter GC 67.88%
#2nd letter GC 46.89%
#3rd letter GC 88.60%

#Codon AA Fraction Frequency Number
GCA    A     0.077     7.772      3
GCC    A     0.462    46.632     18
GCG    A     0.462    46.632     18
GCT    A     0.000     0.000      0
....

There are a number of EMBOSS servers if you want to run the analysis online.

Not online but CodonW does this.

For the web services, you can find some services in the BioCatalogue.

You can then run those services with Taverna.

You could try http://www.bioinformatics.org/sms2/codon_usage.html and their codon usage tool. There is a mechanism to use the tool off-line. See link on the above page. I have no experience with this tool, but you may need to concatenate your individual sequences into one.

http://www.bioinformatics.fr/bioinformatics.php?subsection=Codon%20usage

CodonW can concatenate genes to one sequence and then calculates the overall codon usage offline on Windows.

Leaving this here for the future reference... Following up on the concatenation approach, you could make use of the Bioconductor packages in R to do this: concatenate the sequences using Biostrings, and then analyse codon usage with coRdon.

https://bioconductor.org/packages/release/bioc/html/Biostrings.html

https://bioconductor.org/packages/release/bioc/html/coRdon.html