Image tools

Compress a PDF or image towards a 2MB limit

If an application asks you to compress a PDF to 2MB, check the completed document rather than judging it by the number of pages. This tool attempts a strict ceiling of 2,097,152 bytes, or 2 MiB. A website using decimal 2MB allows 2,000,000 bytes instead. Supported images can use the same target, but PDFs and images follow different compression methods and need different checks.

Local tool available

What this tool does

Compress an image or PDF to below 2 MiB where possible.

Loading the file tool…

Find the exact requirement for this document

Read the guidance beside the upload field and any linked instructions. Confirm whether the limit applies to each file or to the whole group of evidence. Also check the accepted formats, any page-count restriction and whether the organisation expects the complete document.

Do not rely on a rule copied from a different portal or a different year of the same application. If the site does not state a clear maximum, ask the organisation rather than guessing that 2MB is universal. The target on this page is a processing option; it does not establish any particular service's current policy.

Understand why the PDF is large

A PDF can contain real text, vector drawings, photographs, scanned pages and embedded resources. Its visible page count does not reveal how those parts are stored. A one-page scan from a camera can be larger than a much longer document exported as text.

If you still have the source document, inspect it for oversized photographs or duplicate content and export a fresh PDF where appropriate. If the source is paper, review the scanning settings and the clarity of the capture. Reducing a large image at source can address a different problem from compacting the PDF container around it.

What this target actually tries

Select one supported file and press Process files. For a PDF, the tool rewrites the document using a compact storage method and checks whether the new file is strictly below the target. It does not rasterise text or downsample embedded scan images as part of this attempt.

That distinction sets realistic expectations. A PDF with avoidable structural overhead may become smaller. A scan-heavy PDF whose images already dominate its size may not. If the lossless result remains above the ceiling, the tool reports an unreachable target and does not label that oversized document as ready.

If your source is an image, the tool can re-encode it and reduce its dimensions. Those changes can affect visual quality. The image route is therefore not equivalent to the lossless PDF route, even though both appear on the same fixed-target page.

Interpret the result in bytes

A result below 2,097,152 bytes has met the tool's binary ceiling. It has not necessarily met a decimal 2,000,000-byte rule. A file of 2,050,000 bytes illustrates the gap: it is within the first threshold and outside the second.

Measure the downloaded copy and compare it with the destination's stated convention or exact number. Leave some room below the cap where possible. A rounded label of 2.0MB can hide the relevant difference, and changing the filename to include under-2mb does not alter its contents.

If the original file is already below the tool's ceiling, it may be returned unchanged. Read that note carefully. Repeating the same run cannot be relied on to move an already-small file further below a stricter decimal maximum.

Review the complete application document

After downloading, compare the page count with the original and inspect every required section. Check names, dates, signatures, totals and reference numbers. Pay particular attention to pages with pale stamps, small tables or coloured annotations that carry meaning.

For a document that originally had selectable text, try selecting a short passage in the output. If it contains form controls or other special features, test the ones the recipient needs. The tool checks structure and size; it cannot certify that every document-level feature satisfies the organisation's requirements.

Keep the signed original when digital signature validation matters. A newly saved file can be different even if its pages look the same. Ask the receiving service how it wants a signed document supplied when the original exceeds the upload cap.

When lossless compression cannot get there

Return to the best available source instead of repeatedly compressing the same PDF. A fresh export from a document application may remove unnecessary image weight while retaining actual text. A scan can sometimes be recreated at a more suitable resolution, provided the result remains complete and legible.

Do not remove required pages or substitute a low-resolution screenshot of the first page. If the organisation allows separate evidence files, splitting or extracting permitted sections may help, but check whether its limit is per file or combined. If it requires one complete document, request an approved alternative when the necessary information cannot reasonably fit.

An unreachable target is not a judgement about the document's validity. It means the available method cannot provide a verified file under this ceiling. The right next step depends on the source and the receiving rules.

For images, protect dimensions as well as detail

Supported JPG, PNG, WebP and decodable HEIC images can be attempted. Check the format of the output, because a destination may accept only certain image types. An already-small file can retain its original format instead of being converted automatically.

Image dimensions may decrease during compression. Verify the final width and height against any minimums. If transparency would be lost by producing JPG, choose an appropriate explicit background or keep the refusal. View the saved result on the intended background and inspect the smallest meaningful features.

Keep the new file separate from the old attachment

Save the output with a clear name and attach that exact download to the form. Remove an old selected attachment if the interface requires it. When a preview is available, confirm that it shows the intended version and all required pages.

If rejection continues, read the current message. Format restrictions, encryption, expired sessions and service errors are different from excess bytes. Use the file compatibility checker for the rules you can supply, and contact the receiving service when a file meets those rules but is still refused.

Input limits and local processing

The source must fit the 25 MiB input limit and the applicable image-pixel or PDF-page safeguards. Encrypted PDFs are not supported. A large decoded image or complex PDF can fail safely before a usable output is produced.

The selected content is processed in this browser and is not sent to a Make My Files Work server for compression. Your original remains intact. Keep it until you have checked the downloaded copy and completed the submission, so you can return to a reliable version if any issue appears.

Questions that come up

Does this promise a 2MB PDF without any changes?

No. The PDF method avoids intentional image downsampling and page rasterisation, but it still rewrites the file and may fail to achieve the target. Special document features need review, and image inputs can undergo quality and dimension changes.

Can I use a 2MiB result on a site that says 2MB?

Only after checking what that site means and measuring the output. The decimal limit is lower. A result comfortably below two million bytes avoids this particular unit difference, but other upload rules still apply.

Will a ZIP make a PDF acceptable?

Not if the field requires a PDF and rejects archives. A ZIP is a different file type. Follow the receiving instructions rather than using packaging to bypass its stated requirements.