Instagram Photo Downloader

What resolution does Instagram actually serve?

Almost every search for an Instagram photo downloader is really a question about size. Someone has a photo on screen, it looks sharp in the app, and the copy they end up with is soft, small, or somehow both. The gap between those two things is worth understanding, because once you know where the pixels go, everything else on this page follows from it.

The size you see in the app

What your phone shows you is not a file, it is a rendering. Instagram picks an image that suits your screen and your connection, then draws it at whatever size the layout allows. On a modern phone the feed column is around 400 to 450 points wide, and the screen packs two or three physical pixels into every point, so the app is painting roughly a thousand pixels of width into a space the size of a postcard. It reads as very sharp because the pixel density is high, not because the file behind it is enormous.

On the web the arithmetic runs the other way. Instagram's browser layout draws feed photos at a few hundred pixels wide, and on a standard desktop monitor those are literal pixels. A screenshot of that view captures exactly what was on screen, which is why desktop screenshots of Instagram photos come out visibly softer than the same photo saved properly. The screenshot is honest — it captured what the browser had. It just never had the larger version.

The size the server holds

Behind each post Instagram keeps a set of renditions rather than one image. There is a large one, then progressively smaller ones for thumbnails, grid views, previews and slow connections. The large rendition is the one worth having, and for a standard feed post it is 1080 pixels on its longest relevant edge: 1080 by 1080 for a square, 1080 by 1350 for the tall 4:5 crop, and 1080 by 566 for the wide 1.91:1 crop.

The smaller renditions are the ones that show up when a save goes wrong. If you have ever ended up with a 320-pixel square when you expected a full photo, you collected a thumbnail — the version Instagram serves to a grid tile, which is a perfectly good image for a grid tile and a disappointing one on your desktop.

Why 1080 pixels is the ceiling

Instagram resizes on upload. Somebody can shoot a 6000-pixel-wide photo on a full-frame camera, export it at maximum quality, and post it; what the platform then stores for public delivery is a 1080-wide version of it. Every viewer, on every device, in the app or on the web, is looking at that resized copy.

This is the single most useful fact on this page. When people talk about downloading an Instagram photo "in original quality", the original that is available is the largest public rendition, which is the 1080 version. The photographer's export exists, and it exists on the photographer's hard drive. What travels through Instagram is the copy the platform made.

So the realistic goal for any save is: get the largest rendition the platform serves, in one step, without a screenshot in the middle. That is achievable every time for a public post, and everything below is about how.

Where an Instagram photo actually lives

A post has two different addresses and confusing them causes most of the odd behaviour people run into.

The first is the permalink — the address in the browser bar, the one the Copy Link button gives you. It looks like instagram.com/p/ABC123xyz/. That shortcode is the durable handle for the post. It stays valid as long as the post is up, it can be shared, bookmarked and pasted anywhere, and it is what any legitimate tool wants from you.

The second is the image URL on Instagram's content delivery network. It is long, it lives on a hostname full of numbers, and it carries a set of query parameters at the end that look like noise. Those parameters are a signature and an expiry. The CDN checks them, confirms the link was issued recently and legitimately, and serves the bytes.

That expiry is why a trick people remember from years ago no longer works reliably. Copying an image address out of the browser inspector gives you a URL that is valid right now and stops being valid within hours. Save that link in a document and come back next week and it resolves to an error, even though the photo is still perfectly visible on Instagram. Nothing was deleted; the signature simply aged out.

The practical takeaway: keep permalinks, not image URLs. A permalink is a name for the post. An image URL is a temporary ticket to one copy of one file.

How to save a photo from a public post

The flow is the same on every device, and it comes down to getting the right link.

In the app. Open the post so it fills the screen. Tap the three dots in the top right corner and choose Copy Link. On some builds this sits behind Share, then Copy Link. Either way the permalink is now on your clipboard.

On the web. Click the post so it opens as its own page, then copy the address out of the browser bar. Copying while the post is a tile in a profile grid gives you the profile address instead, which is a very common way to end up looking at the wrong thing.

From a story or a share sheet. Links passed around in DMs and stories often arrive with extra tracking text attached — an igshid parameter, a utm_source, sometimes both. They are harmless. Anything that understands Instagram links reads the shortcode and ignores the rest.

A note on the path segment, because it matters more than people expect. A photo post sits under /p/. A Reel sits under /reel/. Older video posts sit under /tv/. If the link you copied says /reel/, what is behind it is a video, and the still frame you can see is a cover image generated from it rather than a photograph. That is a different job with different results, which is why photo and video tools are kept separate.

Once the link is pasted, the useful thing to check before saving is the dimensions of what you are being offered. A number starting with 1080 is the full rendition. A number in the low hundreds is a thumbnail, and it is worth going back for the larger one.

Carousels are different, and this is where people get stuck

If there is one thing that turns a two-second task into a frustrating afternoon, it is a carousel. Someone pastes the link to a ten-photo post, receives one photo, and reasonably concludes that something is broken. Nothing is broken — a carousel just works differently from a single post, and the difference is structural.

Every slide is its own file

A carousel is not one wide image that gets cropped as you swipe. It is a container: one post record holding a list of children, each of which is a complete, independent media object with its own dimensions, its own renditions and its own address on the CDN. Ten photos in a carousel is ten separate files that happen to share a caption, a like count and a permalink.

This is also why the slides can differ from each other in ways that look like glitches. One slide can be a square and the next a 4:5 portrait, because each child carries its own aspect ratio. One can be a video while the rest are photos, which is why a carousel sometimes plays sound when you swipe onto the third slide. Each child was processed on its own terms when the post was uploaded.

The index in the URL

Because the permalink names the container rather than any particular child, a link on its own points at the whole post, and the natural default is the first slide. That is the entire explanation for "I pasted the link and only got one photo".

Instagram gives you a way to be specific. When you swipe to a slide on the web, the address gains a parameter: ?img_index=3 for the third slide, and so on, counting from one. Copy the link while sitting on the slide you want and the index travels with it. Tools that read the parameter will honour it, and people you send the link to will land on the same slide you were looking at, which is useful entirely on its own.

On mobile the parameter is not added automatically, so a link copied from the app points at the container regardless of which slide is on screen. Opening the same post in a mobile browser and swiping there gets you the indexed version.

Getting all of the slides

When the whole set is what you want, there are two sensible routes. The first is to work slide by slide with indexed links — reliable, obvious, and fine for three or four images. The second is to use something that reads the container and enumerates its children, which returns the full list in one pass and is the only pleasant way to handle a twenty-slide post.

Two details worth knowing when you do collect a whole carousel. Filenames come from the CDN and look like long strings of digits, so a set of ten arrives in your downloads folder with names that tell you nothing — renaming them in order as you go takes five seconds and saves a lot of squinting later. And the slides will not all be the same size in bytes even when they share dimensions, because compression depends on the content of each frame. A photo of a plain sky compresses down to a fraction of a photo of a crowd. That variation is normal and is not a sign that one slide came out worse than another.

What this page does

This page reads the public link you paste. No password is ever requested and there is nothing to install.

What that means in practice: you give it a permalink, it asks Instagram's public endpoints what that shortcode refers to, and it shows you what the platform serves for it — the media type, the dimensions of the largest available rendition, and, for a carousel, the list of children behind the container. You then choose what to keep.

Everything it works with is the same information any visitor to that post can see. It does not sign in as you, it does not act on your behalf, and it has no relationship with the account whose post you pasted. That is a deliberate boundary rather than a limitation: a tool that needs your credentials to read a public page has asked for far more than the job requires, and the job here genuinely does not require it.

What happens with private accounts

Worth being direct about, because it is the second most common question after resolution.

When an account is set to private, Instagram serves its media only to approved followers. The check happens on Instagram's side, against the session of whoever is asking, before any image is sent. A request that is not carrying an approved follower's session receives the login wall — which is the correct behaviour, and it is the same behaviour whether the request comes from a browser, an app, or a tool.

So the honest answer is that access to a private account's photos is granted by the account holder, through the follow request, and it is granted to a person rather than to a piece of software. Any service claiming a way around that is describing something the platform does not offer. What those pages typically do offer, once you look, is a survey, an app install or a subscription, with the account name collected on the way in.

There is one distinction that catches people out. A public account can have a private-looking presence — no posts, restricted tags, a locked story — while its profile picture and any posts it has made remain public. Privacy on Instagram is set at the account level for posts, and the setting is what decides delivery. If the posts are visible to you while signed out, they are public; if the profile shows the "This account is private" panel, the account holder has chosen who sees them.

What about profile pictures?

Profile pictures are a related but genuinely separate problem, and treating them as the same thing is why people end up with a blurry 150-pixel avatar.

An avatar is not a post. It lives on the account record rather than in the feed, it is stored as its own small set of renditions, and the app deliberately shows it tiny — a circle a few dozen pixels across in a header, a smaller one in a comment thread. Tapping it opens the profile rather than the image, which is why there is no obvious way to view one at full size from inside Instagram.

The renditions that exist for an avatar are also smaller than the ones for a feed photo. Where a post tops out at 1080, a profile picture is typically stored at a few hundred pixels on its longest edge, because that is all the interface ever needs. Expecting a poster-sized image from an avatar is the mismatch behind most of the disappointment there.

Because the mechanics differ, we keep the two apart: the profile picture viewer handles avatars, explains what sizes are actually stored and how to open one at the largest version the platform holds. This page stays with feed posts and carousels, where the 1080 ceiling and the container structure are the things that matter.

What Instagram does to a photo when you upload it

Understanding the upload pipeline explains the quality of everything that comes out of it, and it also tells you what to do differently if you are the one posting.

The resize step

Instagram normalises dimensions first. Whatever you send, it produces renditions at the widths it serves, with the largest at 1080. Send a 1080-wide file and the resize is a no-op. Send a 4000-wide file and it gets scaled down by a factor of nearly four in a single pass, which is where a lot of fine detail — hair, fabric texture, small text, foliage — turns into mush.

Aspect ratio is normalised too. Instagram accepts a range from roughly 1.91:1 landscape to 4:5 portrait, and anything outside that gets cropped or padded to fit. A very tall image posted to the feed loses its top and bottom, which is a surprise if you composed it carefully.

The re-encode step

After resizing, the image is compressed to JPEG at a quality setting Instagram chooses. Since the source was almost certainly already a JPEG, this is a second lossy compression on top of a first — generational loss, the same effect that degrades a photocopy of a photocopy.

It shows up most in three places: flat gradients like a clear sky get faint banding, sharp edges between contrasting colours get a halo of artefacts, and heavily saturated reds smear slightly. Those artefacts are baked into the served file. Anyone who downloads that photo, by any method, is downloading the compressed version, because that is the version the platform has.

This is worth knowing because it sets a realistic expectation. A saved Instagram photo is a second-generation copy at 1080 pixels. That is genuinely good enough for reference, mood boards, resharing, a phone wallpaper or a small print. It is not the file to hand a printer for a large canvas, and knowing that in advance saves an argument with the printer.

Uploading so that less is lost

If you post photos, a few habits keep more of the original intact.

Export at 1080 wide yourself. Resize in software you control, using a good resampling method, and Instagram's resize becomes a no-op. Your downscale is almost always better than the automatic one because you can sharpen appropriately afterwards.

Use the tall crop. A 1080 by 1350 portrait carries about a quarter more pixels than a 1080 square, and it occupies more of the screen in the feed. It is the highest-resolution shape the platform offers.

Export at high JPEG quality. Give the compressor a clean source. Feeding it a file that is already artefacted means the artefacts get compressed along with the picture.

Sharpen a little, for the screen. Downscaling softens an image. A light sharpening pass after the resize restores the edge definition that the resize took out.

Upload over a good connection. The apps have historically been more aggressive with compression when bandwidth is poor. Posting on wifi rather than one bar of mobile data is a free improvement.

Can the account owner tell?

No notification exists for this, and it is useful to understand why rather than to take it on faith.

Viewing a photo and saving a photo are the same operation at the network level. Your device asks the CDN for an image and the CDN sends it. Whether those bytes are painted on screen and forgotten, or written to a file, is decided entirely on your device, after delivery, in software the platform has no visibility into. There is no round trip back to Instagram that says what happened next.

Instagram does have features that report interaction, and they work because they are interactions: story views are recorded because viewing a story is an event sent to the server, and likes, comments, shares and saves are all deliberate actions with server-side records. A download is not one of those. It is what happens after the delivery of something already sent.

Two related cases are worth separating out. Screenshots of disappearing photos and videos sent in DMs do generate a notification, and always have — that is a specific feature for a specific message type. And a story you view is recorded in the story's viewer list, so opening a story to save something from it does leave a trace, not because of the save but because of the view. Ordinary feed posts, carousels included, sit outside both of those.

What the downloader sites are collecting

Since a search for this brings up dozens of near-identical pages, it is worth knowing what the business behind them looks like, because it explains their design choices.

The pattern is consistent: a single input field, an enormous button, and a page dense with advertising. The service is genuinely free to use, and the revenue comes from three places.

Advertising. The volume of searches here is large and the visits are cheap to serve, so a page that shows several ad units per download is a real business. This is the honest end of the range and it is why so many of these pages exist.

Redirects and interstitials. The second click goes somewhere unexpected: a new tab, a countdown, a "your file is ready, click here" page that is itself an advertisement. Each layer is another impression.

Installs and subscriptions. The end of the chain on some sites is a browser extension, a desktop utility or a subscription with the price below the fold. An extension that can read the pages you visit is a large permission to hand over for a task that a link and a click already accomplish.

The signal worth watching for is a request to sign in. Reading a public post requires the public link and nothing else, so a login prompt on a downloader page is asking for something the task does not need. The same goes for a page that wants a browser notification permission before it will show a result — that permission is for sending you notifications later, not for producing your photo.

None of this makes the whole category suspect. It makes the good ones easy to identify: paste a link, get a result, no account, no install, no chain of countdown pages.

Who owns the photo

Worth a clear-eyed paragraph, because "it is public so it is free" is the assumption that gets people into trouble.

The person who took the photograph owns the copyright in it from the moment it is taken, in essentially every jurisdiction. Posting it to Instagram does not transfer that to anyone — it grants Instagram a broad licence to display and distribute it, which is how the platform can show it to you at all, but the copyright stays with the photographer. A public setting controls who can see the work. It says nothing about what anyone may do with it afterwards.

What that means in practice, in plain terms:

Saving for yourself is the easy case. Keeping a copy for reference, for a private mood board, as a phone wallpaper, or to send to one friend, is personal use, and it is what the overwhelming majority of people are doing. It is the kind of use copyright law was never built to police.

Republishing is the case that needs permission. Reposting to your own account, putting it on a website, using it in a video, or including it in anything commercial is publication. Credit is courteous and it is not a licence — "photo by @name" does not create permission that was not granted. The reliable route is a message asking, and most photographers say yes to a polite request that explains where it is going.

Commercial use raises the stakes considerably. Advertising, packaging, merchandise or anything selling a product brings in both copyright and, if a recognisable person appears, their own rights over the use of their likeness. This is the territory where a written permission is worth having.

Instagram's terms bind you as a user. Separately from copyright, the platform's terms describe what you agreed to when you signed up, including how its data may be collected. Automated bulk collection is treated differently from a person saving a photo, and the distinction is roughly the one common sense would draw.

The short version: saving is normally fine, republishing is a conversation with the person who made it, and money changes the calculation.

Saving without leaving the app

Sometimes the goal is not a file at all, and Instagram's own tools are quicker for those cases.

Bookmark it. The flag icon under a post saves it to your private Saved collection, visible only to you and organisable into named collections. If you are keeping references, this is better than a folder of files — the caption and the account come with it, so in three months you will still know whose photo it was.

Share it to yourself. Sending a post to yourself in a DM keeps it in a thread you can search, and the link stays live.

Screenshot with intent. A screenshot is limited to your screen's pixels, which is why it is not the route to a good file. It is, however, instant and perfect for a detail, a comment thread or a piece of a caption. On a phone with a dense screen a full-screen shot of a single photo is often around a thousand pixels wide, which is close to the served rendition; on a desktop browser it is much less.

Use the platform's own export. If the photo is yours, Instagram's Download Your Information tool produces an archive of everything you have posted, at the resolution the platform stored. Request it in the settings and it arrives as a link by email, usually within a day or so. For getting your own back catalogue out, this is the right tool by a wide margin.

When a saved image comes out small

A short diagnostic list, since this is the complaint behind most repeat searches.

You have a thumbnail. Dimensions in the low hundreds mean a grid or preview rendition was collected instead of the full one. Go back and take the largest option offered.

The source was a Reel. A link under /reel/ holds video. The frame you saved is a cover image, generated by the platform, and it is not a photograph at photo quality. That is a video job.

You screenshotted a desktop browser. The browser layout draws feed photos small, so the screenshot is small. The same post saved from its link gives the full rendition.

The post is old. Instagram's maximum served width has increased over the years — posts from the earlier era were stored at 640 pixels, and some from before that at 612. Those are the files that exist for those posts. A 2014 photo will not produce a 1080 rendition, because one was never made.

The uploader posted a small file. If someone shares a screenshot of a screenshot, the pipeline faithfully preserves exactly that. Nothing downstream can restore detail that was never in the upload.

An upscaler seemed to help. Enlarging software invents plausible pixels; it does not recover the original ones. For a wallpaper this can look genuinely good. For anything where accuracy matters — text, faces, product detail — the invented pixels are invention.

Where the file lands on each device

Small practical differences that account for a lot of "the download did not work" reports.

iPhone and iPad. Safari sends downloads to the Downloads folder in the Files app, not to Photos. To get an image into your camera roll, either use the share sheet and choose Save Image, or long-press the image and pick Add to Photos. This one step is behind most iOS confusion — the file did arrive, it is simply in Files.

Android. Downloads go to the Download folder and the gallery usually picks them up within a moment via the media scanner. If a new image is not showing in your gallery, opening the Files app and tapping it once generally prompts the scan.

Desktop. The browser's usual downloads folder. If your browser is set to ask where to save each file, you get a dialogue; if not, it lands silently and the downloads shelf is the fastest way to find it.

Filenames. CDN files are named with long numeric identifiers. Renaming as you save costs a second and makes a folder of ten carousel slides usable rather than a puzzle.

Common questions

Can I download from a private account? Private posts are served to approved followers, and the check runs against the viewer's own session on Instagram's side. Access comes from the account holder accepting a follow request.

Will they know? Feed posts and carousels generate no notification for saving or viewing. Disappearing photos in DMs are the exception, where screenshots have always been reported to the sender.

Why did I only get the first photo? The post is a carousel, and the plain permalink names the container. Copy the link while on the slide you want in a mobile or desktop browser, so the img_index parameter travels with it, or use something that enumerates the whole set.

Can I get the true original the photographer shot? The original as uploaded stays with the photographer. Instagram distributes its own 1080 rendition, and that rendition is the best copy available to anyone through the platform. For anything larger, asking the photographer is the route that works.

Do I need an account? Not to read a public post. Instagram does show a login prompt to signed-out browsers after a few page views, which is a browsing limit rather than a privacy setting.

Is a browser extension a better route? Extensions are convenient and they run on every page you visit, which is a broad permission for a narrow task. A link-based tool does the same job with a much smaller footprint.

Can I download several posts at once? Working through links one at a time is what the public endpoints support comfortably. Heavy automated collection is what Instagram's rate limits are designed to notice, and it is treated quite differently from a person saving a photo.

What about the caption and the comments? Those are text on the post page and can be copied straight from it. They are not part of the image file, so a saved photo arrives with no caption attached — worth noting if you are archiving something where the caption is the point.

Does saving a photo affect the post's numbers? Instagram counts likes, comments, shares and bookmarks. Fetching an image is not among the counted interactions, so the post's statistics stay as they were.

The short version

Instagram serves a resized copy of every photo, capped at 1080 pixels on its longest served edge, re-encoded once on upload. That copy is the best version available through the platform, and getting it cleanly is a matter of using the post's permalink rather than a screenshot. Carousels hold several independent images behind one link, which is why a plain permalink returns the first slide and an indexed one returns the slide you meant. Private accounts serve their posts to approved followers, decided by the account holder. Avatars follow their own rules and are covered on the profile picture page. And the photographer keeps the copyright throughout, so saving for yourself is ordinary and republishing is a conversation.