92 MB LibreOffice to fit inside AWS Lambda compressed with Brotli

Usage no npm install needed!

<script type="module">
  import shelfAwsLambdaLibreoffice from 'https://cdn.skypack.dev/@shelf/aws-lambda-libreoffice';



85 MB LibreOffice to fit inside AWS Lambda compressed with Brotli

Inspired by chrome-aws-lambda

:information_source: Compiled LibreOffice version:


$ yarn add @shelf/aws-lambda-libreoffice

NOTE: Since version 2.0.0 npm package no longer ships the 85 MB LibreOffice but relies upon libreoffice-lambda-layer instead. Follow the instructions on how to add a lambda layer in that repo.


const {convertTo, canBeConvertedToPDF} = require('@shelf/aws-lambda-libreoffice');

module.exports.handler = async () => {
  // assuming there is a document.docx file inside /tmp dir
  // original file will be deleted afterwards

  if (!canBeConvertedToPDF('document.docx')) {
    return false;

  return convertTo('document.docx', 'pdf'); // returns /tmp/document.pdf

Or if you want more control:

const {unpack, defaultArgs} = require('@shelf/aws-lambda-libreoffice');

await unpack(); // default path /tmp/instdir/program/soffice.bin

  `/tmp/instdir/program/soffice.bin ${defaultArgs.join(
    ' '
  )} --convert-to pdf file.docx --outdir /tmp`


  • Please allocate at least 1536 MB of RAM for your Lambda function.
  • It works only in Amazon Linux 2, so it won't work locally on Linux or macOS. However, you could run it in Docker using lambci/lambda:nodejs12.x image
  • If some file fails to be converted to PDF, try converting it to PDF on your computer first. This might be an issue with LibreOffice itself

See Also


Smoke test that it works: ./test.sh. Make sure to clone libreoffice-lambda-layer repo alongside


$ git checkout master
$ yarn version
$ yarn publish
$ git push origin master


MIT © Shelf