Image Info - Check the Format, Size and Quality of Any Image

Drop an image and see what the file actually is: its real format, its dimensions, its colour depth and channels, whether it has transparency or animation, how it is compressed, and whether it is intact. Nothing is uploaded. The bytes are read in your browser and the file never leaves your device.

Every value here is read from the bytes, never from the file name. If a file is named .jpg but is really a PNG, this page says so and explains why it matters. A renamed file is usually harmless.

Drag & Drop your images here

or

From my Computer
By URL
From Google Drive
From Dropbox

Import from your computer, a public URL, Google Drive, or Dropbox.

JPG, PNG, WebP, GIF, TIFF, BMP, ICO, SVG, HEIC, AVIF - up to 100 MB per file

🔒 100% processed on your device ⚡ No upload, no network calls 🗑 Files are never stored on a server
Options

Summary shows format, dimensions and colour. Full adds encoding, animation, which metadata blocks are present, and the file integrity checks. Switching this re-renders what is already on screen and never re-reads your files.

Declared only reads the header flags, which is instant. Scan pixels decodes the image and counts how many pixels are actually see-through, which is the only way to find an alpha channel that is carrying nothing. It is slow on very large images and is not available for HEIC or AVIF.

The container headers, straight from the raw bytes: the format signature, the frame headers, the colour and compression fields, the animation blocks and the end markers. Nothing here is changed, re-encoded or uploaded.

Metadata is summarised here, not listed. For the full tag list use the EXIF viewer, to strip it use the EXIF remover, and for print size at every resolution use the DPI checker.

The estimated JPEG quality is worked out from the compression tables inside the file. JPEG does not store the quality setting anyone typed, and different encoders reach the same visual result with different numbers, so treat it as a close approximation rather than the original value.

A file whose extension does not match its real format is usually just a renamed file and is normally harmless. It matters because some upload forms and older software go by the extension and will reject or mishandle it.

An alpha channel does not mean anything is actually transparent. Turn on the pixel scan to find out, and expect it to be slow on very large images.

Frame delays are what the file asks for. Browsers clamp very short delays, so an animation often plays slower than its stated duration.

The compression ratio compares the file against uncompressed pixel data at the same bit depth. It is a rough reference point, not a quality score.

The colour profile name is read from the profile description inside the file. The profile itself is not validated.

Bytes after the end marker are usually padding added by a phone or a delivery network. They can also be an appended file. The count is reported without any claim about what they are.

SVG is a text format, so it has no fixed pixel size and can scale to anything. An SVG can also contain scripts, which is why many sites will not accept SVG uploads.

HEIC and AVIF are read from the container rather than decoded, so the pixel scan is not available for them.

Everything is processed locally in your browser. Nothing is uploaded to our servers.

This image info tool answers the question you ask before every other question: what exactly is this file? It reads the real format from the first bytes rather than trusting the extension, then reports the pixel dimensions, the colour model and bit depth, the chroma subsampling, whether the file is animated, whether it declares transparency, an estimate of the JPEG quality, and whether the file is actually intact. Everything is read in your browser and nothing is uploaded. It is the broad first look in the image metadata tools group, next to the EXIF viewer, the DPI checker and the EXIF remover.

Two things here are hard to find anywhere else. It tells you when a file's name is lying: a PNG saved as photo.jpg is identified as a PNG and flagged, which matters because plenty of upload forms and older software go by the extension and will reject it. And it runs a file integrity pass, checking for a missing end marker, failed chunk checksums and bytes appended after the file should have finished, then says "No problems found" out loud when everything passes, because that is a result people came for. Metadata is summarised here rather than listed, so for the full tag list use the EXIF viewer and for print size at every resolution use the DPI checker. The rest of the image tools sit alongside, including the format converters and the compression tools.

10
Formats read by magic bytes
7
Report sections per file
0
Files written or uploaded
100 MB
Per file, batch supported
🏷️

It Catches a Lying Filename

The format comes from the first bytes, never the extension. A PNG named photo.jpg is identified correctly and the mismatch is flagged, calmly, with what it actually means for an upload form.

🧩

A Real Integrity Check

Missing end markers, failed chunk checksums and bytes appended after the file should have ended. When everything passes it says so, rather than leaving the section empty.

🎨

Colour Detail Nobody Shows

Chroma subsampling, bit depth per channel, colour model and the embedded profile name, plus an estimated JPEG quality worked out from the compression tables.

👁️

Real Transparency, Not Declared

An alpha channel does not mean anything is see-through. Switch on the pixel scan and it counts what is actually transparent, which finds channels carrying nothing at all.

How to Check What an Image File Really Is

Four steps, and there is no button to press for the main answer. Files are read the moment they arrive, straight from the container headers.

📤

1. Add the File

Drag it in, browse, or import from a URL, Google Drive or Dropbox. Ten formats are recognised, including ICO and SVG, up to 100 MB each.

2. Read the Six Tiles

Format, dimensions, megapixels, aspect ratio, colour and file size answer most questions in one glance, before you read anything else.

📋

3. Switch to Full Detail

Full adds encoding, animation, which metadata blocks are present and the integrity checks. Summary keeps it to format, dimensions and colour.

💾

4. Scan, Copy or Export

Turn on the pixel scan if transparency matters, then copy the report or download it as a text file covering every image you loaded.

Check an Image File Now

What This Image Info Tool Reads, Format by Format

There is no before-and-after picture here because nothing is produced. What changes from one format to the next is where the answers live and how much a container is willing to tell you about itself. Every row below is a code path in the tool on this page, not a line from a specification.

Format Dimensions from Colour detail Animation Integrity checks Notes
JPEG The frame header, SOF Bit depth, colour model, chroma subsampling, estimated quality Not possible in JPEG End marker, segment walk, trailing bytes Baseline or progressive is reported, and quality is always labelled an estimate
PNG IHDR Bit depth, colour type, palette size, interlace method APNG frame count, delays and loop IEND, per-chunk checksums, chunk lengths The only format here with a real checksum per chunk, so corruption is named
WebP VP8X canvas, else VP8 or VP8L Lossy, lossless, alpha flag ANIM and ANMF frame count and loop RIFF size against the buffer, chunk walk Extended WebP declares its features in a flag byte, which is read directly
GIF Logical screen descriptor Global and local colour table sizes Frame count, total duration, loop count 0x3B trailer, trailing bytes Delays are what the file asks for. Browsers clamp very short ones
TIFF, BMP, ICO IFD0, the BITMAPINFOHEADER, the icon directory Bits per sample, samples per pixel, palette Not applicable Header structure and offsets An ICO lists every size in its directory rather than only the first
SVG, HEIC, AVIF Width, height and viewBox, or the ispe box Vector, or the container's own colour boxes AVIF sequences are reported Parsed with DOMParser, or the box walk SVG scripts are counted and never run. HEIC and AVIF are read, not decoded

The format itself is always identified from the file's opening bytes, so every row above applies to the real contents rather than to whatever the filename claims. Metadata is summarised, never listed: for the full tag list use the EXIF viewer, for print resolution use the DPI checker, and to strip what you find use the EXIF remover.

Image Info: What Is This File, Really?

Almost every image problem starts with an unanswered question about the file itself. The upload form rejects it and will not say why. The photo looks flat and nobody can explain it. The download might be complete or might have stopped halfway. Someone insists it is a JPEG and the software disagrees. This page answers all of those from the bytes, before anyone has to guess.

The first thing it does is refuse to trust the filename. Formats are identified from their opening signature, so a PNG saved as photo.jpg is reported as a PNG, and the mismatch is flagged. That is usually harmless, because someone typed the wrong extension or a phone did, but it matters in practice: plenty of upload forms and older software route on the extension and will reject or mishandle a file whose name lies about it. The Library of Congress digital format descriptions are a good reference for how precisely each of these containers is actually specified, which is why reading the signature is the only reliable way to know what you are holding.

The second thing worth knowing is the deliberate boundary. This page summarises which metadata blocks are present and does not list them, because that is the EXIF viewer's job. It gives a single line about DPI and leaves the print size table to the DPI checker. Four narrow tools that each answer one question well beat one page that answers four questions vaguely, and the whole image metadata tools group is built that way on purpose.

⚖️How This Compares to Other Ways to Inspect a File

Method Detects a renamed file What it misses
This image info tool Yes Nothing on the container. Metadata is summarised by design.
Right click, Properties No Trusts the extension, and shows no colour or encoding detail.
Opening it in an image editor Partly Tells you it opened. Says nothing about subsampling or integrity.
Browser developer tools No Reports the served MIME type, which can be wrong too.
The Unix file command Yes One line of output, no integrity pass, no print or colour maths.
Desktop ExifTool command line Yes Nothing much, and it is an install plus a terminal to learn.
Guessing from the thumbnail No A thumbnail renders happily from a file that is subtly broken.

The last row is not a joke. A truncated PNG frequently renders its top half and looks fine in a file browser, and a JPEG with four kilobytes appended after its end marker displays perfectly while breaking strict validators. Neither problem is visible by looking. The integrity section exists precisely for the files that look fine and are not, and it is the section people come back for. If what you need is to change the file rather than understand it, that is the format converters, the image resizer or the compression tools.


🔍When Checking the File Details Is the Right Move

An Upload Keeps Getting Rejected

Forms reject on format, dimensions or file weight, and rarely tell you which. Reading all three at once, plus whether the extension matches the real contents, usually identifies the problem in seconds.

🎨

The Colours Look Wrong Somewhere

An embedded profile that one app honours and another ignores is the usual cause. The report names the profile, the colour model and the bit depth, which is enough to work out which end is at fault.

🧩

You Suspect the File Is Damaged

Missing end markers, failed chunk checksums and truncated chunks are all caught. A file that renders in a preview can still be incomplete, and this is how you find out before it matters.

📦

A PNG Is Far Bigger Than It Should Be

Colour type, palette size and whether the alpha channel is carrying anything explain most oversized PNGs. Turn on the pixel scan and an entirely opaque alpha channel shows itself immediately.


🚩What This Page Deliberately Leaves to Its Siblings

🔍

The Full Tag List

You get a presence summary: EXIF yes or no, XMP, IPTC, colour profile. For every camera, lens, exposure and GPS field, that is the EXIF viewer, which is built for exactly that.

📏

Print Size and DPI Conflicts

You get one line about the DPI tag. The full print size ladder, and the cases where a file carries two or three resolution values that disagree, belong to the DPI checker.

🧹

Removing Anything

This page writes no file at all. When the report shows metadata you would rather not share, the EXIF remover strips it without re-encoding the picture.

🔄

Changing the File

Nothing here converts, resizes or compresses. Once you know what the file is, the converters, the resizer and the resize to KB tool take over.


What This Tool Does That Others Do Not

🏷️

Extension Mismatch, Explained Calmly

A renamed file is flagged with what it actually means rather than an alarming warning. It is usually harmless, and it is also the reason a strict upload form is rejecting you.

"No Problems Found" Is Printed

A clean integrity result is a real answer, so the section says so out loud instead of quietly showing nothing. Silence is indistinguishable from a check that never ran.

🖼️

Chroma Subsampling Is Reported

4:2:0 against 4:4:4 explains a great deal of "why does this JPEG look soft", and almost nothing else in a browser will tell you which one you have.

🔢

The Quality Figure Says Estimated

JPEG stores no quality setting. The figure comes from the compression tables, and the word estimated appears every single time the number does, because a confident wrong number is worse than none.


What the Image Info Tool Actually Reports

  • Identifies JPEG, PNG, WebP, GIF, TIFF, BMP, ICO, SVG, HEIC and AVIF from their opening bytes, and flags any file whose extension disagrees.
  • Reports pixel dimensions, megapixels, aspect ratio as a reduced ratio, and orientation, from the container's own frame header.
  • Reports the colour model, bit depth per channel, palette size where indexed, and JPEG chroma subsampling with its sampling factors.
  • Estimates JPEG quality from the quantisation tables against the standard reference tables, labelled as an estimate every time it appears.
  • Reports baseline against progressive JPEG, PNG interlacing, and whether a WebP is lossy or lossless.
  • Detects animation in GIF, APNG, animated WebP and AVIF sequences, with frame counts, delays and loop behaviour.
  • Reports declared transparency from the header, and optionally scans the pixels to count what is genuinely see-through.
  • Runs a file integrity pass: end markers, per-chunk checksums on PNG, truncated chunks, RIFF size agreement and bytes appended after the file ends.
  • Lists every size inside an ICO directory, and parses SVG as text with DOMParser, counting scripts and foreignObject elements without ever running them.
  • Handles batches, exports one combined plain text report, and makes no network call after the page loads.

🔢How to Use the Image Info Tool

  1. Add one file or several Drag them in, or import from your computer, a URL, Google Drive or Dropbox. Ten formats are recognised including ICO and SVG, up to 100 MB each, and each file gets its own result.
  2. Read the six tiles first Format, dimensions, megapixels, aspect ratio, colour and file size. If the format tile disagrees with the filename, that alone often explains whatever brought you here.
  3. Switch to Full when you need the rest Full adds encoding, animation, the metadata presence summary and the integrity checks. Summary keeps it to format, dimensions and colour. Switching re-renders what is on screen and never re-reads the file.
  4. Scan pixels only when transparency matters This is the one control that does new work. It decodes the image and counts genuinely transparent pixels, which is slow on very large files and unavailable for HEIC and AVIF. Then copy or download the report.

⚙️Technical Notes and Honest Limits

The figure that needs the most care is the estimated JPEG quality. JPEG does not store the quality setting anybody typed. What it stores is a pair of quantisation tables, and the estimate here works backwards from those against the standard reference tables. Different encoders reach the same visual result with different numbers, so treat it as a close approximation of the setting used, not a recovered value. The word estimated appears every time the figure does, in the report and in the exported text file, and that is deliberate.

🔄 Why an alpha channel is not the same as transparency

A PNG saved as colour type 6 carries a full alpha channel whether or not a single pixel uses it, and that channel costs roughly a quarter of the file. Declared transparency, which is what the header says, is instant to read. Whether anything is actually see-through can only be answered by decoding and counting, which is what the pixel scan does. The scan reads in horizontal strips rather than allocating the whole image at once, and it deliberately does not downscale first, because interpolation invents partial transparency at the edges and would report a number that is simply wrong.

💾 What bytes after the end marker actually mean

Every one of these formats has a defined ending: a JPEG end of image marker, a PNG IEND chunk, a GIF trailer byte. Data occurring after that point is reported by count and nothing more. Usually it is harmless padding added by a phone, a camera or a content delivery network, and the file works perfectly everywhere. Occasionally it is a second file appended to the first. The count is the only thing that can honestly be said about it, so the count is all the tool says, without any accusation attached.

🗎 Why SVG is treated as text and never rendered

SVG is XML, not pixels, so it is parsed with DOMParser and the resulting document is never attached to the page. Nothing inside it can execute. The report counts script elements, inline event attributes and foreignObject elements, and states the width, height and viewBox, noting when the size is relative rather than intrinsic. That script count is the honest reason so many sites refuse SVG uploads: a script inside an SVG runs when the file is opened directly in a browser. Reporting it is information, not an accusation about your file.

Why an animation plays slower than the report says

The frame delays shown are the values written in the file. Browsers clamp very short delays, so a GIF asking for ten milliseconds a frame is usually played at closer to a hundred, and the real running time comes out several times longer than the stated duration. The tool reports the file's own numbers because those are the facts about the file, and says plainly that the played duration will often differ. The same applies to loop counts: zero means forever, and it is printed as forever rather than as zero times.


💡Real Situations This Page Handles

01

A form keeps rejecting an image that looks perfectly normal. The format tile shows it is a PNG despite the .jpg name, and converting it properly with PNG to JPG makes the extension true and the upload succeed.

02

A developer wonders why a PNG asset is three times the size of the JPEG beside it. Colour type 6 with a fully opaque alpha channel, found by the pixel scan, explains it and points straight at the fix.

03

A download may have been interrupted. The integrity section reports a missing IEND chunk and a truncated final chunk, which settles it without opening the file in anything that might make matters worse.

04

A photographer cannot see why one export looks softer than another at the same dimensions. One is 4:2:0 and the other 4:4:4, which the colour tile reports directly and nothing else in a browser will tell them.


FAQs - Image Info

Drop it in and the answer appears immediately. The format is identified from the opening bytes of the file, which is the only reliable method, rather than from the extension on the end of the name. If the two disagree the tool says so directly, naming the real format and the one the filename claims. Ten formats are recognised, including ICO and SVG.

Usually not. Someone typed the wrong extension, or an app saved it that way, and the picture is perfectly fine. It matters in one specific situation: plenty of upload forms and older software route files by their extension, so a name that lies about the contents gets rejected or mishandled. Converting it properly makes the extension true rather than just renaming it again.

JPEG stores brightness at full detail and colour at reduced detail, because eyes notice brightness far more. At 4:4:4 the colour channels keep full resolution. At 4:2:0 they are stored at a quarter, which shrinks the file and softens fine coloured edges, especially red text on a dark background. It explains a great deal of why one export looks softer than another at identical dimensions.

It is a close approximation, and the word estimated appears every time the number does. JPEG does not store the quality setting anyone typed. What it stores is the compression tables, and the estimate works backwards from those against the standard reference tables. Different encoders reach the same visual result with different numbers, so treat it as the setting most likely used rather than a recovered value.

It runs a real integrity pass. It checks for the format's end marker, walks the internal structure for truncated segments or chunks, verifies the per-chunk checksums that PNG carries, and reports any bytes sitting after the file should have finished. When every check passes it prints 'No problems found', because a clean result is an answer rather than an empty section.

Every one of these formats has a defined ending. Data after that point is reported by count and nothing more. Most of the time it is harmless padding added by a phone, a camera or a content delivery network, and the file works everywhere. Occasionally it is a second file appended to the first. The count is all that can honestly be said, so that is all the tool says.

Because a declared alpha channel and actual transparency are different things. A PNG saved as colour type 6 carries a full alpha channel whether or not a single pixel uses it, and that channel costs roughly a quarter of the file size. Switch the transparency option to scan pixels and the tool decodes the image and counts what is genuinely see-through, which finds channels carrying nothing.

Only as a summary. You get whether EXIF, XMP, IPTC and a colour profile are present, and a single line about the DPI tag. The full tag list belongs to the EXIF viewer and the print size table belongs to the DPI checker, both of which are linked from the results. Four narrow tools each answering one question well beats one page answering four vaguely.

Yes. SVG is XML rather than pixels, so it is parsed with DOMParser and the resulting document is never attached to the page, meaning nothing inside it can execute. The report counts script elements, inline event attributes and foreignObject elements and states the dimensions. That script count is also the honest reason many sites refuse SVG uploads in the first place.

The delays shown are the values written in the file, which are the facts about the file. Browsers clamp very short frame delays, so an animation asking for ten milliseconds a frame typically plays at closer to a hundred and runs several times longer than its stated duration. Loop counts are reported the same way, with zero printed as forever rather than as zero times.

JPEG, PNG, WebP, GIF, TIFF, BMP, ICO, SVG, HEIC and AVIF, each read from whichever headers that container uses. An ICO lists every size inside its directory rather than just the first. HEIC and AVIF are read from the container rather than decoded, so their details are reported but the pixel transparency scan is not available for them.

- WHO USES THIS -

Built for Everyone in India

Anyone holding a file that will not behave: an upload that keeps failing, a photo that looks wrong, a download that may or may not be complete, or a format nobody can quite identify.

Web and App Developers Designers and Studios QA and Support Teams Students Filling Online Forms Photographers Print and Prepress Archivists and Librarians Anyone Debugging a Bad Upload

Find Out Exactly What Your File Is

Real format, dimensions, colour depth, animation, transparency and whether the file is intact. Ten formats, read from the raw bytes in your browser, with nothing uploaded and nothing changed.

Open the Image Info Tool