Resize image in KB.
Made beautifully simple.

Read the upload requirements before changing your image

Copy the destination’s instructions into a short checklist. Look for the accepted format, maximum and minimum file size, width, height, aspect ratio and any photography rules. “Below 100 KB” is different from “between 50 KB and 100 KB.” A file that is smaller than the maximum may still fail a minimum-size rule. Likewise, a 50 KB photograph can be rejected because it is the wrong width or because the form accepts only JPEG.

Separate technical requirements from visual requirements. File size and dimensions are properties this tool can help you adjust. A plain background, visible facial features, an uncropped signature or a recently taken photograph require you to inspect the image itself. Compressing a photo cannot establish that it meets an authority’s identity checks. If the service asks you to take a live photograph inside its application, a resized file is not a substitute for that step.

Check the current form or the issuing organization’s own notice. A preset is a convenient way to enter starting settings, but instructions can differ between applications with similar names. Our document photo preparation guide explains that distinction. Keep a copy of the original outside the browser, especially before working with a personal photograph or a document you might need again. Avoid editing your only copy through a series of downloads and recompressions.

How to resize an image in KB, step by step

1. Choose your source image. Use Browse Files, drag a supported image into the upload area, or paste an image from the clipboard where your browser allows it. Choose the original photograph or scan when possible. A picture saved from a messaging app may already have lost detail. After opening the image, check that it is upright and that the file shown is the one you intended to use.
2. Enter the requested maximum. Set Target file size and select KB or MB. Quick targets provide convenient starting values such as 20 KB, 50 KB, 100 KB and 200 KB. Enter a different value if your destination specifies one. The tool searches for an output within that limit; it does not simply rename the original file or promise that every image can fit every target.
3. Set dimensions when they are required. Open More options to use Dimensions & units, Crop & Rotate and the framing controls. Enter the destination’s width and height. Custom dimensions stay fixed during compression. Without custom dimensions, the tool may make the image smaller in pixels to try to meet the size target. Always check the output dimensions after it finishes.
4. Select an accepted output format. Choose JPG, PNG or WebP. For a form requiring JPG, select JPG even if the source was PNG. A transparent logo usually needs PNG in this tool. The most compact format is not useful when the receiving website does not accept it. Inspect any background fill when converting an image with transparent areas.
5. Review, download and reopen. Compare the preview, actual file size and pixel dimensions. Look for an above-limit warning. Download the image when the result is suitable, then open the saved copy before uploading it elsewhere. Keep the original and use a descriptive filename so you can distinguish the final version from earlier attempts.

If your task has no hard size cap, turn off Target file size and use the manual Image Quality control for JPG or WebP. This lets you compare a few quality settings instead of forcing an unnecessarily small file. The percentage is an encoder setting, not a measurement of how much visual information remains. PNG does not use that adjustable quality parameter in this tool.

KB, MB and pixels measure different things

KB and MB describe the amount of data in the saved file. Pixels describe the image’s width and height. Two images with identical dimensions can have very different file sizes: a simple white background is easier to encode compactly than a detailed scene with grass, hair, shadows and camera noise. That is why a fixed width does not predict a fixed KB value, and why a preset cannot promise the same result for every source.

Reducing dimensions removes pixels. Reducing JPG or WebP quality changes how the remaining pixels are encoded. Cropping removes part of the picture. Each approach changes a different part of the result. A portrait with useful detail throughout may benefit from moderate compression; a signature surrounded by empty paper may benefit more from careful cropping. Choose the adjustment that removes unnecessary information while preserving what the reader or receiving service needs.

This tool calculates 1 KB as 1,024 bytes and 1 MB as 1,048,576 bytes. Some destinations use decimal units instead. For example, a limit of 100,000 bytes is slightly smaller than the tool’s 100 KB setting of 102,400 bytes. Where a portal states an exact byte limit, use that limit as the reference and leave some margin. Size displays are rounded, so a value close to the boundary needs extra care.

Choose a target that fits the job

Use the largest file that comfortably satisfies the destination when preserving detail matters. A 200 KB allowance does not mean you should squeeze the image down to 20 KB. If there is no stated limit, consider its display size and content instead of choosing a tiny number by habit. These targets are examples of how to think about a size budget; they are not specifications for a particular exam, passport or application.

A maximum is not an instruction to fill every available byte. If a clear result is comfortably below the allowed limit, there is usually no benefit in making it larger merely to display the target number. However, a destination may also require a minimum file size. Check both boundaries and try a higher target or different settings when the output is below a required minimum.

Choose JPG, PNG or WebP for the actual destination

JPG for photographs and forms that require JPEG

JPG and JPEG name the same format. In this tool, JPEG output supports adjustable compression quality and a solid background. Start with it for a photograph when the destination accepts JPEG. Check features that matter to the task, such as eyes in a portrait or writing on a photographed label. If the input has transparency, inspect the selected background fill in the output. See the JPG compression guide for that workflow.

PNG when transparent areas or clean graphics matter

The PNG option is useful for transparent artwork and graphics where you want to avoid lossy encoding. PNG encoding is lossless, but resizing and cropping still change the source image. This tool does not provide a PNG quality slider or palette optimizer. To meet a difficult cap, it can reduce pixel dimensions when custom dimensions are not set. With fixed dimensions, it may return a file above the requested limit and show a warning. Check the PNG size guide before choosing a very small cap.

WebP for destinations that accept it

The WebP option also uses adjustable encoding quality here. It is worth comparing with JPG for an image intended for your own website, provided that your publishing system accepts it. This tool’s WebP path fills a solid background, so choose PNG when you need to retain transparency. For application forms, check accepted types first. Use the WebP resizing guide for format-specific help. Changing a filename from .webp to .jpg does not convert the image; select the correct output and download a newly encoded file.

Set dimensions without distorting the subject

Keep Ratio Locked when the destination only specifies one dimension or allows the original proportions. This helps avoid stretching a person’s face or making a circular logo oval. If exact width and height imply a different shape from the original, decide how to handle the difference before exporting. Unlocking the ratio does not create missing content; it lets you enter a different shape that the framing mode must then fill.

Cover / Crop fills the output frame while keeping the subject’s proportions, but parts of the image can fall outside the edges. Fit & Pad keeps the whole image inside the frame and adds space around it when the shapes differ. Stretch fills the dimensions by changing proportions. Use the preview to inspect each result rather than assuming the selected numbers guarantee suitable framing. Padding and cropping may be unacceptable for a particular application.

For example, fitting a wide landscape into a square can leave bands above and below it, while covering a square cuts off the sides. A portrait might lose hair or shoulders under a tight crop. Use Crop & Rotate to prepare the framing deliberately, then recheck the output. Our pixel dimension guide focuses on this choice. Enlarging a small source adds pixels, but cannot restore detail that the original never contained.

Convert centimeters, inches and DPI carefully

Digital upload instructions usually work most clearly in pixels. When you have physical measurements, the tool uses DPI to calculate a pixel width and height. The conversion is: pixels = centimeters ÷ 2.54 × DPI. An illustrative 3 × 4 cm image at 300 DPI converts to approximately 354 × 472 pixels after rounding. This is a mathematical example, not an official document photo specification.

The DPI field is a conversion input. It does not write your selected print-DPI metadata into the downloaded image. If a printing workflow checks a resolution tag or a physical print size, use suitable printing software to set and verify those properties. An image can have enough pixels for a print while a separate program still needs the correct print settings. See the centimeter conversion guide for more examples.

Changing DPI while using centimeters changes the calculated number of pixels. It is therefore not a magic compression switch. Fewer pixels may reduce file size, but can also remove useful detail. If your form supplies exact pixel dimensions, enter those directly instead of guessing a DPI. If it only specifies a printed photograph, check whether it actually accepts a digital upload and whether digital alteration is allowed before using an online resizer.

Judge quality by what someone needs to see

Look at the preview at a useful size and inspect the downloaded copy. A small display can hide blocky edges, soft writing or damage to fine lines. For a photograph, check the face, boundaries and any important label. For a signature, follow the thinnest strokes and make sure every end remains visible. For a screenshot, try reading the smallest text that the recipient will need.

When a result looks poor, start again from the original. Recompressing a previously compressed JPEG adds another encoding step and may make damage harder to fix. Try a less restrictive target if the destination permits it. Remove irrelevant borders, use the required dimensions and compare accepted formats. If every part of a large image is important, a tiny allowance may simply be incompatible with a readable result.

The tool’s comparison is useful for spotting compression and framing changes, but inspect the original separately when crop decisions matter. The displayed reference can share the current framing, so it is not proof that every original edge has survived. No quality percentage can certify legibility or photo compliance. Keep the version that meets the actual task, and retain your source for a different destination later.

Examples you can adapt to your own image

These examples describe settings and decisions, not measured compression results. Your output depends on the source image, its detail and the browser encoder. The stated dimensions and file-size allowances below are illustrative and must be replaced with your destination’s instructions.

A portrait for a form with a 100 KB maximum

Suppose a form accepts JPG and asks for a 400 × 500 pixel image below 100 KB. Open the original portrait, enter 400 and 500 as custom pixel dimensions, and use the preview to choose suitable framing. Select JPG and set the target slightly below the maximum if you need margin for differing size units. Check the face and the full downloaded image. If the file remains oversized, keep the required dimensions and consider a clearer, simpler source photo that still meets the photography rules. Do not quietly shrink below a mandatory minimum width.

A signature surrounded by empty paper

Suppose the destination permits a cropped scan and accepts a 20 KB JPG. Crop unnecessary paper while keeping every part of the signature and an appropriate margin. Apply the required dimensions if provided, then try the size limit. Thin strokes disappearing are a reason to revise the scan or settings. Repeatedly lowering quality on a faint image can make it unusable. The signature preparation guide explains the checks; verify current requirements separately for a named institution.

A transparent logo for a website

Start with the original PNG and choose dimensions appropriate for its intended display. Keep PNG output and inspect the transparent edges and the selected framing/background options. If the file remains larger than your preferred budget at fixed dimensions, decide whether smaller dimensions are acceptable or whether you need a dedicated PNG optimizer. Switching to JPG may reduce size while adding a solid background; that is a different visual result and should be a deliberate choice.

Photos to attach to an email

Use a larger allowance when the recipient needs to inspect detail. Prepare each image, then check the combined attachment size against your email service’s limit. A 1 MB target applies to each exported image, not the entire message. If you create a ZIP, verify its total size and that the recipient can open archives. Keep descriptive filenames and avoid sending a low-resolution preview when the recipient requested original evidence or full-resolution photographs.

Prepare multiple images and check them individually

Select several supported files or use Add More in the workspace. Set the shared target and output format, review the images, and download a batch ZIP when ready. Shared custom dimensions can be useful for a set of matching thumbnails, but they may crop a mixture of portrait and landscape photographs differently. Inspect important edges on each image instead of checking only the first preview.

The file-size target applies to each image. The ZIP can be larger than that target because it contains multiple results and archive overhead. Already compressed image formats may gain little additional reduction from ZIP packaging. Extract the archive and check the files your destination actually accepts; an upload form asking for a JPG generally expects the individual image, not an archive.

If the tool warns that some batch outputs exceed the limit, revisit those images before submitting them. One easy-to-compress file does not prove that the rest fit. Large batches also consume browser memory, particularly when high-resolution sources are decoded. Work in smaller groups if your device slows down, keep the tab open during processing, and save the results before resetting the workspace or refreshing the page.

Why an image may still fail to upload

The result is above the requested maximum

First check the actual file size and the warning. A detailed image at fixed dimensions may not fit a very small budget, especially as PNG. If accepted, compare JPG or WebP. If dimensions are flexible, reduce them or remove irrelevant borders. If the destination requires both strict dimensions and a small cap, preserve the dimensions and reconsider the source. Increasing the target only helps when the destination allows that larger size.

The portal says “invalid format”

Read its accepted types and export a new file in the required format. Renaming a PNG as JPG does not change its contents. Select the downloaded result rather than the original from your Downloads folder. A damaged image or a file the browser cannot decode may need to be reopened and exported by an image editor before it can be processed here.

The size appears correct but is rejected

Check for minimum size, pixel dimensions and an aspect-ratio rule. Leave margin below a maximum rather than relying on a rounded display at the boundary. Confirm that KB and MB were not confused. If all visible technical requirements are satisfied, the portal may be checking a separate photo rule or may have a service problem. Consult its instructions and support channel; do not assume every rejection is a compression error.

The image is blurry or the writing is broken

Return to the original and compare less aggressive settings. A photograph of a faint signature may require a sharper scan, better lighting or a closer capture. Enlarging blurred text makes the blur larger. Avoid sharpening, retouching or other changes that an official application forbids. The useful test is whether the required detail remains clear in the actual saved image.

The browser cannot open the source file

The upload control accepts JPG, PNG, WebP, GIF and BMP. It does not accept PDF, HEIC, RAW camera files or SVG as inputs. Export a supported image from appropriate software, ensuring that the conversion does not violate the destination’s rules. If a supported file still fails, try a fresh copy, a current browser or a smaller source. Record the format, browser and error when reporting a problem; personal photos are not needed for an initial report.

The result has missing edges or a solid background

Review Cover / Crop, Fit & Pad and Stretch in More options. The first can remove edges; the second can add surrounding space; the third can distort proportions. Check the separate crop selection too. For transparent content, choose PNG and inspect the background settings. This tool does not remove an existing background or reconstruct a subject that was missing from the source.

What browser processing means for your images

The resizing workflow reads the file on your device, draws the image in the browser and creates a downloadable result. It does not upload the selected image files to our servers for compression. You can use the tool without creating an account. Keep your own copies because the workspace is not a cloud backup and does not provide an account where you can recover an earlier image.

Website delivery and optional analytics are separate from image processing. Hosting providers receive network information needed to serve the page, and analytics can measure site use when permitted. The Privacy Policy explains these services and the available choices. You can reopen Privacy choices in the footer. Changing an analytics choice may reload the page, so download any work you want to keep first.

Browser export generally does not retain the original camera’s EXIF metadata, but do not treat that as a complete privacy audit. Visible addresses, document numbers, faces and background details remain part of the image unless you change the framing. Inspect what the recipient can see before sharing. For help, describe the technical issue without emailing identity documents or sensitive photographs.

What this resizer can and cannot do

The image resizer changes dimensions, framing, output format and encoding settings. It exports a separate image and supports batch ZIP packaging. It does not preserve animated GIFs as animations, restore missing detail or certify identity photographs. For documents, use Image to PDF to arrange images on pages, or PDF Size Reducer for an existing PDF. Each document tool explains its own quality and feature limitations. GIF input produces a still image.

Compression uses the browser’s image encoder. The size search has practical limits, and a result may remain above the target. Custom dimensions stay fixed even in that case. This is useful when a destination requires an exact pixel frame: you see that the size requirement remains unresolved instead of receiving a smaller image with unexpected dimensions.

Outputs can vary between browsers and source images. We do not promise a universal reduction percentage, an exact byte count or unchanged quality for lossy formats. An official photo process may prohibit digital edits or require a professional provider or live capture. Follow that process when it applies. A convenient tool should help you meet the actual instructions, not encourage you to bypass them.

A final checklist before uploading

Open the downloaded image and work through this list. Checking the saved file catches common mistakes such as selecting an old version, exporting the wrong format or overlooking a crop that removes part of the subject.

• Correct destination: you used the current instructions for this specific form, account or publishing system.
• Accepted format: the downloaded image is JPG, PNG or WebP as required, rather than a renamed source file.
• Allowed size range: the actual file is below the maximum and above any minimum, with margin where size units differ.
• Required dimensions: width, height and proportions match the stated requirement; check them again if automatic downscaling was used.
• Readable subject: the face, signature, writing or product detail remains clear at the intended viewing size.
• Suitable framing: important edges remain visible, and any padding or background is permitted.
• Correct file selected: the upload dialog points to the final saved result, while your original remains safely stored.

After submission, check the destination’s own preview or confirmation. A successful local download proves that a file was created; it does not prove the receiving service accepted it. If you need to revise the result, return to your original image and change the setting responsible for the problem. This keeps the work focused and avoids unnecessary rounds of compression.