open the tool

download an instagram reel or post

paste the link to a public reel or post and get the file. you don't need an account or an extension.

how it works

  1. paste the link into the box on the tool.
  2. pick video or audio, and a quality if you don't want the default.
  3. grab resolves the link and the file arrives.

usually that last step is just a direct link to the platform's cdn. the data never touches the grab server. some sites won't serve a client they don't recognize and those get tunneled through the server and then back out. no writes to disk happen. youtube and some others hand over the video and the audio separately but your browser stitches those back together via ffmpeg compiled to webassembly.

quality walks down a ladder: 1080p, then 720p, 480p, 360p, 240p if the higher ones are not there. audio mode gives you mp3.

what you get

instagram keeps changing its mind about what a logged-out request can see. that's usually why a link stops working. it normally comes back after a while.

if the format is wrong

there's a remux tab that runs ffmpeg inside your browser so the file stays on your machine. it converts between mp4, webm, mp3, ogg, opus, wav, m4a.

"auto" rewrites the container and copies the codec across untouched. converting to mp3 or wav re-encodes instead and you lose a little each time.

what doesn't work

some sites don't work at all:

what grab keeps

you don't need an account to download anything.

nothing you fetch gets written to disk on the server. direct downloads never touch it and tunneled ones just pass through and get dropped. grab doesn't keep a history of what you downloaded and it doesn't run analytics. nothing about you is sold to anyone.

the full privacy policy and terms are in the about tab.

questions

stories?

only while they're public and still up. once it expires, or if the account is private, it's not reachable.

will the account know?

no. nothing logs in so it never counts as a view.

it worked yesterday and fails today

that's usually instagram tightening what a logged-out request can fetch. give it a day before you assume the link is bad.