Choose the required image, not the largest available version
A camera may offer a very large original that is useful for editing or print but unnecessarily heavy for a small web attachment. That does not mean you should immediately make it tiny. Read the required width, height and file type first, then choose a source that can supply those dimensions clearly.
For a product photograph, check the details a buyer or reviewer needs to see. For a screenshot, check the smallest writing. For a document photo, check names and reference numbers. These different uses place different demands on the picture, so there is no single visual-quality setting that is correct for every 1MB task.
What the compression attempt can change
Select one supported file and press Process files. Image processing can adjust encoded quality and reduce pixel dimensions while it searches for a file below the fixed target. It stops when it finds a verified result that fits; it does not promise a mathematically optimal balance of every possible quality and size setting.
JPG, PNG, WebP and decodable HEIC can be processed, but they do not all behave the same way. PNG and WebP generally keep their formats on this path, while HEIC processing can produce JPG. Animated GIF compression is not offered. If a file is already below the target, it can instead be returned with its original bytes.
Do not confuse format conversion with a size result
A website may accept JPG but reject WebP even when both are comfortably under its cap. This target page does not automatically know or apply that website's accepted-format list. Check the resulting format and use a dedicated converter where the receiving instructions require one.
This matters for an unchanged small file. A HEIC picture already below the target can be returned as HEIC, because reducing size was unnecessary. That does not solve a HEIC compatibility rejection. Convert it explicitly to an accepted format, then measure the converted output, which may be larger than the source.
Review the pixel dimensions after processing
A photograph can become smaller partly because the output has fewer pixels. Open the dimensions checker on the downloaded copy and compare its width and height with the destination's minimums. A small byte figure cannot tell you whether a listing photo, identification image or screenshot remains large enough.
Do not enlarge a heavily reduced output simply to restore the numeric dimensions. Upscaling can increase the pixel count without restoring the lost detail. Return to the original and prepare a more appropriate version if the compressed result is too small. Keep the source until the final upload has been checked.
Check photographs and screenshots differently
For a photograph, look at faces, textured surfaces and smooth colour transitions. Compression can make fine texture look smeared or introduce blocks in difficult areas. View the image at a realistic display size as well as zooming into important details.
For a screenshot or graphic, inspect small letters, sharp borders and thin lines. JPG artefacts can be particularly visible around high-contrast edges. PNG may be more suitable for those features when the destination accepts it, although that can make the target harder to reach. Choose based on the actual content rather than a blanket rule that one format is always smaller or better.
If the image has transparency and processing would produce JPG, use the explicit white or black background choice only when it suits the artwork. Otherwise retain the protective refusal. A change to the background can make pale logos or lettering unreadable even when the byte target is met.
Use the exact byte count at the boundary
The built-in binary ceiling is 48,576 bytes above decimal 1MB. For example, a result of 1,020,000 bytes passes the tool's target but is larger than a decimal million-byte limit. This is a unit difference, not evidence that either system has misread the file.
Inspect the saved file's bytes and leave suitable space below the receiving cap when possible. If the form uses a precise maximum, use that number. A filename such as under-1mb does not establish the size of its contents, and a rounded display such as 1.0MB can conceal a value on either side of the limit.
For PDFs, know the method's limits
The PDF route attempts lossless repacking of the document's stored structure. It does not resample the images inside a scan or offer a choice of aggressive compression levels. A text-based PDF with inefficient storage may benefit; an already compact image-heavy file may not.
If the requested PDF cannot fit, revisit the original document export or scan settings. Keep all required pages, readable small print and any necessary colour evidence. Do not convert a searchable document into screenshots solely to force it through a size check when the recipient expects the original document features.
Find the cause if the upload still fails
Attach the downloaded result explicitly and confirm its filename in the form. If an old attachment remains selected, the website may still be rejecting the original. Then check the exact message: size, format and dimensions point to different corrections.
A file can also be rejected because it is protected, damaged, unsuitable for the field or being submitted through an expired session. Use the compatibility checker with the actual rules you have. Once measurable requirements are satisfied, ask the receiving service about a persistent rejection rather than repeatedly reducing the image.
Keep the process manageable on your device
Inputs have a 25 MiB byte cap and additional safeguards for decoded image pixels and PDF pages. A small compressed source can still expand substantially in memory when decoded. Very large or complicated files may therefore be refused even though the desired output is only around 1MB.
Processing remains in the browser on your device, without sending the selected file to a Make My Files Work server. Save the output in a recognisable place, open it independently and keep the source available. If an attempt fails, use a suitable smaller source export or an authorised simpler document rather than assuming the original has been damaged.
Questions that come up
Does this keep the original image dimensions?
Not necessarily. The compression attempt can reduce width and height. Check the output against any minimum dimensions and return to the original if the result has become too small for the receiving field.
Can I use the target for an email attachment?
You can use it when that size is appropriate, but your email service's rules may concern the whole message rather than this file alone. Check the actual provider guidance and account-specific instructions.
Why does the tool return the same small file?
If it already fits the built-in target, altering it is unnecessary. The result note explains that the bytes were preserved. Check separate format and dimension requirements if the upload is still rejected.