TL;DR: No — Tumblr does not strip EXIF metadata from photos the way Instagram, Facebook, or X do. Tumblr's own privacy policy states plainly that it "may collect information describing your camera, camera settings, or EXIF information" from uploaded images, and the platform has shipped a feature since 2011 that displays this data — camera model, aperture, shutter speed, ISO, focal length — directly on photo posts. Independent testing that uploads a photo with embedded GPS coordinates and then downloads it to check what survived has repeatedly found Tumblr among the platforms where location data comes through intact, unlike Instagram or Twitter/X, which re-encode images aggressively enough to drop most metadata as a side effect. Tumblr does resize and JPEG-compress photos for its dashboard feed, but compression is not metadata removal, and the two get conflated constantly. Worse, once a post is reblogged, that copy lives independently on someone else's blog — deleting your original or your entire account does nothing to the metadata sitting in every reblog downstream. The dependable fix is stripping the file before you ever hit post.
Does Tumblr remove photo metadata?
Not reliably, and in some respects not at all by design. Tumblr occupies an unusual spot among major platforms: rather than treating EXIF as a liability to scrub on ingest, it has historically treated it as a feature to expose. When Popular Photography covered Tumblr's 2011 update, the platform had just added the ability to show a photo's camera make, model, aperture, shutter speed, ISO, and focal length right on the post — visible to anyone who viewed it. That's the opposite instinct from a platform trying to protect you from your own metadata.
The GPS field is the one that matters most, and testing bears out that it isn't reliably removed either. A comparison test that uploaded a photo carrying GPS coordinates to a range of platforms and then downloaded each copy to check found Tumblr's file still carrying its location, in the same bucket as Flickr — while Facebook, Instagram, and Twitter stripped it. That result lines up with independent GitHub tooling built specifically around this gap: projects like tumblr-exif-check exist because researchers found enough Tumblr users posting geotagged photos to build a scraper around it, and separate research tracked the share of geotagged photos on the platform climbing from roughly 1% in 2012 to 13% by 2016. None of that happens on a platform that strips coordinates on upload.
What actually happens when you upload a photo to Tumblr
Tumblr does process every image you post, just not in the way people assume. The platform runs uploads through an automated resizing pipeline: an image destined for the main dashboard feed gets downsampled to a fairly small width for the feed view, and even the "high-resolution" version you reach by clicking through tops out well below your original file's dimensions. Tumblr also converts most uploads to JPEG regardless of the source format, applying its own compression tuned for file size over fidelity.
None of that is a metadata-removal step. Resizing changes pixel dimensions. Re-encoding to JPEG changes the container format. Neither operation is designed to touch the EXIF APP1 segment sitting in the file, and whether metadata survives a given re-encode depends entirely on which image library performed it and how that library was configured — something Tumblr has never documented publicly. What we do know, from Tumblr's own privacy policy and from repeated third-party testing, is that camera data and GPS coordinates have survived the trip often enough to be a documented, recurring finding rather than a one-off bug. If you want the deeper technical breakdown of what a JPEG's metadata containers actually hold, our explainer on what EXIF data is covers the fields platforms like this tend to leave in place.
Tumblr's Neue Post Format doesn't change the picture
Since 2018, Tumblr's backend has represented every post — photo, text, video, or otherwise — as a structured JSON document under what it calls the Neue Post Format (NPF), replacing the older HTML-blob post types. NPF gives every image its own media block with dimensions, a type, and a URL to Tumblr's CDN copy of the file. That's an architectural change in how Tumblr stores and serves posts, not a privacy change in how it handles the bytes inside an uploaded photo. The image still goes through the same resize-and-recompress pipeline before it's referenced from an NPF media block, and nothing about representing a post as structured JSON instead of HTML implies anything gets scrubbed from the file along the way.
The reblog problem: metadata that outlives your deletions
Reblogging is the feature that makes Tumblr's metadata handling genuinely different from a platform like Instagram or Reddit. When someone reblogs your photo post, Tumblr doesn't create a live reference back to your original file — it makes an independent copy that now lives on the reblogger's own blog, complete with whatever the original upload contained. Tumblr's terms are explicit about what that means for you: if you delete your post, your blog, or your entire account, any reblog of that photo remains exactly as it was, on whatever blog reblogged it, with the same metadata that was in the version you originally posted.
Photo by Alexey Demidov on Pexels.
This is a fundamentally different failure mode from the one we cover in posts about messaging apps, where the risk is a single recipient getting a copy of your file. On Tumblr, a popular post can accumulate hundreds or thousands of reblogs within hours, each one a fully independent snapshot of the image — GPS coordinates and all, if they were present. You cannot claw that back later. Cleaning your original post after the fact does nothing for copies that already propagated, which is the same structural problem we describe in more general terms in what metadata scrubbing actually removes — stripping is a per-file, per-copy operation, and it can't retroactively apply to files that already exist elsewhere.
How Tumblr compares to other platforms
It helps to place Tumblr against platforms with a similar upload-and-redistribute model. Reddit's image pipeline re-encodes photos aggressively enough on ingest that most EXIF gets dropped as a byproduct, even though privacy was never the design goal — the re-encoding happens for bandwidth and consistency reasons, and metadata removal just rides along. Instagram and Twitter/X behave similarly: their processing pipelines are aggressive enough that GPS rarely survives, even without an explicit "strip metadata" step anywhere in their public documentation.
Tumblr sits apart from that group. Its resize pipeline is real, but multiple independent tests have found it insufficient to reliably remove GPS, and the platform's own EXIF-display feature suggests metadata removal was never a design priority the way it clearly was for Reddit's or Instagram's ingest pipelines. That distinction matters if you're choosing where to post something you'd rather not have geotagged — the platform's general approach to image processing is not a substitute for checking, and in Tumblr's case testing suggests you generally shouldn't assume anything gets removed.
What about video posts?
Video on Tumblr carries the same exposure as photos, arguably more. A video file's metadata can include the recording timestamp, device make and model, and GPS coordinates baked in by your camera app if location services were active while filming — and unlike a single photo's GPS tag, a video's location metadata often represents where filming started, which can be a very specific and identifiable moment. Tumblr transcodes uploaded video for its player much as it resizes and recompresses photos, and exactly as with photos, transcoding is not a documented metadata-removal step. If GPS survives Tumblr's photo pipeline in testing, there's no established reason to assume its video pipeline behaves differently, and no public statement from Tumblr claims that it does.
What stripping still doesn't hide
Even after you've cleaned a file, two limits are worth being honest about. First, removal only applies to the copy you actually clean. If you post a stripped photo but the original — GPS and all — still sits in your camera roll or a cloud backup, sharing that original anywhere later exposes exactly what you thought you'd already handled. Cleaning is a per-file operation, not a property that follows an image around once multiple copies exist.
Second, metadata is only the invisible layer. A photo with no GPS tag can still show a street sign, a recognizable storefront, a reflection with a license plate in it, or a skyline that places the shot just as precisely as coordinates would have. Removing the EXIF block closes the machine-readable leak — the one nobody thinks to check — but it says nothing about what's visible in the frame itself. Our post on how GPS coordinates get embedded in photos in the first place walks through where that data originates if you want the fuller picture of what you're working against.
How to remove photo metadata before posting on Tumblr
Since Tumblr doesn't reliably do this for you, and since reblogs can make a bad upload permanent within hours, the safer sequence is to clean the file before it ever reaches your dashboard. Metadata Cleaner does this entirely in your browser: drop in a photo or video, and it rewrites the file without its EXIF block, GPS IFD, XMP and IPTC records, and embedded thumbnail, handing you back a clean copy in about a minute. Because it works on video and audio as well as still images, it closes the same gap whether you're posting a photoset, a clip, or an ask with an attached voice memo. Nothing uploads to us — the processing happens locally, on your machine.
Photo by Leila Abboud on Pexels.
Once you've posted the cleaned copy, it stops mattering whether Tumblr resizes it, recompresses it to JPEG, or reblogs it two thousand times — there's nothing left in the file for any of that activity to expose. For the full walkthrough across desktop and phone, see our guide on how to strip EXIF data from a photo. As the EFF's ongoing work on privacy makes clear, metadata is a whole category of exposure — location, timestamps, device identifiers — and no amount of trusting a platform's resize pipeline substitutes for checking the file yourself.
Tumblr is not hiding what it does here; its own privacy policy tells you it collects camera and EXIF information, and it built a feature specifically to surface that data on your posts. The platform simply never promised to strip it, and the reblog mechanic means a single unclean upload can outlive every attempt you make to take it back. Strip the file before you post it, and none of that machinery has anything left to work with. Try Metadata Cleaner free and post the clean copy.