Sycago
All toolsReport an issue ↗
SYCAGO / IMAGE TOOLS Files stay on your device
LESS FILE SIZE. MORE POSSIBILITIES.

Free image
compressor.

Free online image compressor for JPG, PNG and WebP. Set a KB limit, resize your photos, convert formats, and download a whole batch.

No signupNo watermarkNo image uploads
YOUR IMAGE WORKSPACE

Make room for better images.

Read the guide ↗

Drop your images here

or click to choose files

JPG · PNG · WebP · BMP · AVIF¹
Up to 30 images · 30 MB per file · 150 MB per batch
Or paste an image while this tool is focused.

Image queue 0 / 30

Your images will appear here.
Start with a photo or try the sample above.

Ready when you are. No files selected.

¹ AVIF and BMP input depends on your browser. Animated images are not supported. HEIC, SVG and TIFF are not accepted. Large images may be limited by your device's available memory.

01 / PRIVATE

Your files. Your device.

Compression runs locally in your browser. This widget makes no external network requests.

02 / FLEXIBLE

One image or a whole batch.

Adjust quality, size and dimensions, then download individual files or a ZIP.

03 / CLEAR

See what changed.

Compare images, inspect dimensions and export a size report before you move on.

THE SYCAGO IMAGE GUIDE

Free online image compressor: smaller files, sensible settings

A photo can look excellent on your screen and still be unnecessarily large for its next destination. A website may only display it in a small content column. An application form may reject anything above a particular file-size limit. A colleague may need an email attachment that opens quickly on a phone. Image compression helps you prepare a useful copy for that specific job, while keeping your original available for future editing.

The Sycago image compressor brings several preparation steps into one browser workspace. You can choose a maximum size in KB, adjust JPG or WebP quality, resize dimensions, change output formats, rotate a picture, and process multiple images together. You can also download a ZIP archive and a CSV report. No account, watermark or external image-processing service is required by this widget. Your device performs the work, so its available memory and browser capabilities still matter.

How to compress an image online

  1. Add a still image. Choose files, drag them into the upload area, or paste an image while the tool has focus. The queue shows the original file size and dimensions.
  2. Choose a mode. Use Target file size when you need an upload below a stated limit. Use Quality control when visual appearance is more important than a strict maximum.
  3. Choose your output. Select JPG, PNG, WebP, Smart mode, or a supported original format. Pick resize options if the current pixel dimensions are excessive.
  4. Process and inspect. Click Compress images. Read the result notes, compare the previews, and verify the final dimensions before downloading.
  5. Save your copies. Download images individually, or click Prepare ZIP followed by Save ZIP to collect the completed outputs. Keep your original files separately.

Start with one representative image before applying the same settings to a large group. A setting that suits a landscape photograph may work poorly for a screenshot containing tiny text. When you change any settings, the tool clears previous outputs so that a download cannot silently contain an older configuration. Your original selected files remain in the queue, ready for another processing run.

Compress an image to 20 KB, 50 KB, 100 KB or a custom limit

Target mode is designed for a maximum file size. Enter a number or use a preset, such as 20 KB, 50 KB, 100 KB, 200 KB or 500 KB. The tool first tries to encode the image within your chosen quality ceiling. For JPG and WebP, it can lower quality to find a smaller output. If that still exceeds the target and you allow smaller dimensions, it reduces the pixel dimensions and tries again.

A result below 50 KB is usually more useful than a file padded to exactly 50 KB. This compressor does not add filler bytes simply to make a number look exact. Its result notes tell you whether the encoded file is within the requested limit. If the target cannot be met within the available processing attempts, the output remains clearly marked as exceeding the target. Check that note before submitting it anywhere.

File-size units deserve attention. In this tool, 1 KB means 1,000 bytes and 1 MB means 1,000,000 bytes. Some software uses binary units, where 1 KiB is 1,024 bytes. A strict upload system may display size differently. When an application rejects a file very close to its limit, choose a slightly lower target, inspect the downloaded byte count, and retry. Always follow the receiving service's current instructions for format, dimensions and permitted size.

Practical image target-size considerations
TargetPossible starting useWhat to check
10–30 KBSmall signatures, simple graphics or constrained form uploadsFine strokes, readable text and minimum dimensions
50–100 KBSmall photographs, compact illustrations or thumbnailsFacial detail, colour transitions and required pixel size
200–500 KBWebsite content images and larger previewsDisplay dimensions and visible artefacts
Custom limitA specific workflow or destination requirementFinal bytes, accepted format and visual suitability

These are example starting points, not official requirements for any examination, identity document, recruitment portal or social platform. A very small target can make a complex image unsuitable even when the file technically passes the size check. Meeting the byte limit and producing a useful image are separate decisions. Inspect both before using the output.

JPG, PNG or WebP: choose the format for the image

JPG for photographic content

JPG is a practical choice for photographs and destinations that ask for JPEG uploads. Its quality setting trades some image information for a smaller file. Moderate compression can work well for photographic detail, but strong compression may create blocky edges, colour smearing or ringing around text. JPG does not keep transparency. When a transparent input becomes JPG, the tool fills the background using your selected background colour.

PNG for transparency and crisp graphics

PNG supports transparent pixels and lossless encoding of the rendered image. It can suit diagrams, screenshots and graphics with clear edges. A photographic PNG may be much larger than a comparable JPG or WebP. This tool does not use the quality slider to reduce PNG encoding quality. To make a PNG smaller, reduce its dimensions or choose another format when the destination permits it. Resizing or colour edits still change the rendered image, even though the PNG encoding itself is lossless.

WebP for modern website images

WebP supports compression with a quality setting and can preserve transparency. It is useful when you want to prepare images for a website that accepts the format. A WebP result is not automatically smaller than every JPG or PNG, so compare the actual output size. Some forms and older editing workflows only accept specified file types. Check compatibility before converting an image that another person or system must open.

Smart mode keeps JPG inputs as JPG and normally uses WebP for other supported inputs. If your browser cannot encode WebP, the tool uses its available alternative. The original-format option keeps JPG, PNG or WebP where supported; BMP and AVIF inputs are converted to a supported output. Input support and output support are separate browser capabilities. AVIF export, HEIC conversion and animated-image editing are not part of this compressor.

Reduce pixel dimensions before forcing extreme compression

File size and image dimensions describe different things. File size is the number of bytes stored in the file. Dimensions are the width and height in pixels. A large photograph often contains far more pixels than a small website card needs. Resizing it to a sensible display size can reduce the amount of data that needs encoding and may produce a better-looking result than keeping thousands of extra pixels at a very low quality setting.

Use Limit longest edge when you want a consistent maximum size without manually calculating both sides. A landscape and a portrait image can each fit within the same longest-edge limit while preserving their shapes. Percentage scaling is useful when a batch needs the same proportional reduction. Enlargement is disabled by default because adding pixels does not recover information that was absent in the original. Enable it only when a destination requires larger dimensions and you accept that limitation.

Custom width and height provides three different fitting choices. Fit inside preserves the whole image and returns dimensions that fit within your box. Fill and centre crop produces the requested shape by removing content outside the centre crop. Fit and pad keeps the whole image inside an exact-size canvas, leaving surrounding space where necessary. Padding can remain transparent in PNG or WebP, or use your selected background colour.

Exact dimensions and a very small file limit may conflict. If Target file size mode is allowed to shrink dimensions, it can reduce the final output below your custom width and height. Turn that option off when dimensions are mandatory. The file may then exceed your KB target, which the result notes will explain. For a strict form upload, inspect the final width, height, format and byte count together rather than relying on one preset button.

Balance quality, transparency and colour adjustments

A quality value is a browser encoder setting, not a promise that a particular percentage of visual detail survives. Different images and browsers can produce different file sizes from the same value. In Quality control mode, the selected value is applied directly to JPG or WebP encoding. In Target file size mode, it becomes the maximum quality the tool will try; the final value can be lower when necessary to meet the size limit.

Compare important areas after processing. Faces, small lettering, thin lines and smooth gradients reveal compression problems more easily than large areas of texture. The comparison window scales images to the available screen space, which makes it convenient for overall checks. It is not a substitute for opening the downloaded file at 100% zoom. When text becomes difficult to read, increase the target size, keep more pixels, or try PNG instead of applying stronger photographic compression.

Brightness, contrast, saturation and grayscale are optional adjustments. They apply to every image in the current batch, so use them carefully when files have different lighting or colour requirements. Rotation and flips also apply to the whole batch. If your browser does not support the required colour filter, the tool reports that limitation rather than claiming the edit succeeded. Reset the controls or process the image in a browser that supports the feature.

Transparent edges can look different on a dark website, a white document or a coloured social post. Keep transparency enabled for PNG and WebP when the background should remain open. Disable it to flatten the image against your chosen colour. JPG output always receives a solid background. Check the output against its intended destination, especially when a logo has pale edges that disappear against a similar background.

Practical workflows for websites, forms and social content

Prepare images for a WordPress website

Begin with the dimensions your page layout actually needs. A full-width hero image and a small testimonial portrait have different requirements. Resize first, then compare JPG and WebP outputs at reasonable quality. Use descriptive file names that help your team identify the content. After uploading, also write suitable alternative text in WordPress. Compression does not generate alt text, choose responsive image markup or configure your theme; those decisions belong in the website publishing workflow.

Smaller image files can reduce transfer requirements, but image compression alone does not guarantee a faster site or higher Google rankings. Hosting, scripts, fonts, caching, image loading behaviour and layout decisions also affect the experience. Test a representative page after replacing its images. Confirm that the browser requests the intended files and that the image still looks clear at the sizes visitors actually see.

Prepare an image for an application form

Read the current instructions before editing. A form may specify a minimum resolution, permitted dimensions, background, format and maximum file size. Enter the required dimensions, select an accepted format and choose a conservative KB target. If the form requires a specific crop, ensure that the subject remains correctly positioned; the automatic centre crop cannot judge identification-photo rules. Recheck the downloaded file rather than assuming that every photo with the right byte count will be accepted.

Prepare social media graphics

Use the shape that suits your planned placement, such as a square, portrait or landscape layout. The dimension presets are convenient starting canvases, not a guarantee of current platform recommendations. Fit and pad can help retain a complete design, while centre crop can remove essential words or logos near the edges. Review the entire output before publishing. A social platform may resize and recompress the image again after you upload it.

Share images by email or with a client

For review copies, balance readable detail with a manageable attachment size. A batch ZIP keeps a set together without requiring many individual downloads. Remember that the ZIP archive stores the already encoded images; it is not intended to produce another dramatic reduction in image size. Include the original files separately when a client needs print production, detailed retouching or a future design revision.

Batch image compression and download options

The queue accepts up to 30 images, with a limit of 30 MB per file and 150 MB of selected inputs per batch. These limits help keep browser processing manageable, but they cannot guarantee that every device has enough memory for every permitted batch. Images are processed one at a time. Very large decoded images can still consume much more memory than their compressed file size suggests. On a phone, smaller batches are often the more reliable choice.

You can sort the queue by added order, file name or original size. Remove any item before processing if it does not belong in the batch. During processing, controls are locked to prevent settings and file changes from mixing different configurations. The stop button finishes the current image before stopping the rest of the run. Completed outputs stay available, and each result shows its own status rather than hiding failures behind a batch-wide success message.

Use the optional prefix and suffix controls to distinguish prepared copies from originals. The default suffix is “-sycago”, and you can change or remove it. The ZIP exporter adds a number when output names would otherwise collide inside the archive. The CSV report includes original and output sizes, dimensions, file types, target status and processing notes. It can help a team record what changed without uploading its images to an external reporting service.

Local processing, original files and metadata

The compressor code does not send your selected images to a remote image-processing API. It reads files through your browser and creates downloadable copies locally. There are no external script libraries, web fonts or image services required by the widget. This statement describes the compressor itself. The surrounding website may have its own analytics, advertisements or other scripts, so its broader privacy practices should be described in the website's privacy policy.

Your original disk files are not overwritten by compression. The downloaded output is a separate file, and the browser chooses where to save it. If the option to retain a smaller original is enabled, the tool can copy the original bytes when no edits or conversion are needed and the original already satisfies the relevant target. That avoids making an unchanged image larger, but it also preserves metadata embedded in the original.

Re-encoding through a canvas normally discards much of the original camera or application metadata. This tool does not certify that every piece of sensitive information has been removed. Visible details remain visible, and file names can also contain personal information. If metadata removal is a formal requirement, use a dedicated inspection workflow and verify the resulting file. Clear the queue when finished, and remember that downloads may remain in your browser history or downloads folder.

Troubleshooting image compression results

The output is larger than the input. A conversion, an enlarged canvas or a different encoder can increase file size. An already optimised input may have little room for improvement. Try another output format or reduce the dimensions. The smaller-original option only applies when the output can remain unchanged; it does not ignore a crop, rotation, colour edit or requested conversion just to show a better savings percentage.

The target was not met. Check whether automatic dimension reduction is disabled. If dimensions must stay fixed, increase the permitted file size or accept a lower quality setting where appropriate. PNG can remain relatively large at a fixed size because this tool does not perform palette optimisation. Tiny targets can also be impractical because file headers require space even for a very small image. The result is marked honestly when the requested limit is not achieved.

The image cannot be opened. Check the actual file type rather than changing its extension. A file called “photo.jpg” is not necessarily a JPEG. This tool checks common image signatures before decoding. Unsupported, animated or damaged files are rejected with an explanation. If your browser cannot decode an AVIF or BMP input, first export a still JPG or PNG from an application that can open the source correctly.

The device becomes slow or runs out of memory. Reduce the number of selected files and choose a smaller longest-edge value. The tool limits accepted inputs to 40 megapixels and a 16,000-pixel maximum side, and caps output at 8,000 pixels per side and 40 megapixels. A device can still struggle below those limits. Close unnecessary browser tabs, avoid repeated enormous batches and download successful results before starting another session.

The downloaded image has the wrong appearance. Recheck the resize fit, rotation, flip, transparency and colour settings. A centre crop removes outer content; padding adds space; JPG replaces transparent pixels with a solid background. Canvas processing can change colour profiles and metadata, so use a colour-managed editor for critical print production. This browser utility is designed for convenient web and everyday image preparation, rather than archival or specialist photographic workflows.

Frequently asked questions

Is the Sycago image compressor free?

Yes. The Sycago compressor has no subscription, API key, signup requirement or watermark. It uses the capabilities of the visitor's browser. The file-count and size limits shown in the interface are practical processing limits, not a paid upgrade gate. Normal internet access is needed to load the hosted page initially.

Can I compress an image to exactly 20 KB?

You can set a maximum of 20 KB. The tool aims to produce a file at or below 20,000 bytes rather than padding it to an exact byte count. Quality and, if permitted, dimensions may change. Read the final target status and inspect the downloaded image before submitting it.

Can I compress a photo to 50 KB or 100 KB?

Yes. Select the relevant preset or enter a custom target. The complexity of the photo, chosen dimensions and output format affect the result. Some photos need smaller dimensions to meet those limits. If you must preserve a minimum resolution, disable automatic shrinking and check whether the final file still meets your target.

Will image quality remain completely unchanged?

No general guarantee is possible. JPG and quality-controlled WebP encoding may discard detail. Resizing, cropping and colour adjustments also change pixels. PNG encodes the rendered image losslessly, but rendering can still change the source. Use the preview and inspect the downloaded file at full size to judge whether it is suitable.

Why does the quality slider not reduce my PNG?

Browser PNG export does not use the lossy quality control applied to JPG and WebP. To reduce a PNG with this tool, lower its dimensions or convert it to an accepted alternative format. This implementation does not include a dedicated indexed-colour palette optimiser, so detailed PNG images can remain relatively large.

Can I keep a transparent background?

Yes, for PNG and supported WebP output, with transparency enabled. The checkerboard in previews represents transparent areas. JPG cannot store transparent pixels, so the selected background colour is used instead. A background already painted into an input image cannot be removed simply by enabling transparency.

Does this tool remove backgrounds with AI?

No. It preserves existing transparency or flattens the image against a solid colour. It does not identify a subject or remove an existing background. That keeps its purpose clear and avoids presenting a basic transparency setting as an AI background-removal service.

Can I process multiple images at once?

Yes. Add up to 30 eligible files within the displayed size limits. They use one shared configuration and are processed sequentially. Download the outputs individually, or click Prepare ZIP and then Save ZIP to download the archive. If images require different crops or colour settings, process them in separate batches so each receives appropriate adjustments.

Why are my exact custom dimensions smaller in the output?

Target mode may reduce dimensions when automatic shrinking is enabled. Fit inside also preserves the original aspect ratio, so it may return dimensions smaller than the custom bounding box. To require an exact canvas, choose centre crop or padding, then disable target-driven shrinking and verify the final result.

Does changing dimensions change print DPI?

This tool controls pixel dimensions, not an arbitrary print-DPI metadata setting. Physical print size depends on the pixel count and the resolution used by the printing workflow. Entering a particular width in pixels does not certify a specific printed size. Use your layout or print software to confirm physical dimensions.

Are HEIC, TIFF, SVG or GIF files supported?

No. The accepted inputs are still JPG, PNG, WebP, BMP and browser-supported AVIF. Animated PNG, WebP and AVIF are rejected where their animation markers are detected. Convert other sources to a suitable still image first. The tool does not promise to preserve multiple frames, vector shapes or specialist image data.

Does the tool work on a mobile phone?

It is designed with responsive controls for smaller screens, but processing depends on the phone's browser and available memory. Use a smaller batch when working with high-resolution camera photographs. If a selected format cannot be decoded or exported, follow the displayed error and try a compatible input or output.

Are my original photos changed?

No. Processing creates separate downloadable outputs. The tool does not write over the files you selected from your device. Keep originals in your own storage, especially if you may need them for future edits, print work or another destination with different quality requirements.

Does a smaller image automatically improve SEO?

No. A lighter image may reduce transfer size, but ranking and page performance depend on many factors. Use relevant images, suitable alternative text, appropriate dimensions and a useful page experience. A compressor is one part of website preparation, not a promise of rankings, traffic or a particular performance score.

Why did the tool retain my original file?

With the smaller-original option enabled, an unchanged original may be retained when it is no larger than the encoded result and meets the relevant size limit. Its processing note makes that clear. A retained original keeps its embedded metadata. Disable the option if you want the image re-encoded even when the output grows.

Does the ZIP add another layer of image compression?

The ZIP exporter packages the processed image files together using stored entries. It does not re-encode them or reduce their visual quality again. Because the images are already compressed, the archive can be slightly larger than their combined file sizes due to file names and archive headers.

Can I use the tool without an API or a server?

The image operations do not require a backend or external API service. The standalone HTML file can run in a compatible browser, and the embedded version can run within a website page that permits its script. Hosting the page and any surrounding website services remain separate from image processing.

Why do results differ between browsers?

Browsers can use different image encoders, colour handling and implementations of the same quality parameter. A value of 85 does not guarantee identical bytes everywhere. For strict submissions, check the actual downloaded file on the device you are using rather than relying on a previous result from another browser.

What information is included in the CSV report?

The report lists original and output file names, byte sizes, dimensions, output format, size reduction, target status, whether an original was retained, and processing notes. It is generated on your device. Review file names before sharing the report, since they may reveal information about the images or their subjects.

How should I choose the best settings?

Begin with the destination's requirements. Choose sensible pixel dimensions, select a compatible format, and use a moderate quality ceiling. Inspect the result, then adjust only what is necessary. A slightly larger file with readable text or clear facial detail is often more useful than the smallest technically achievable download.

For browser behaviour and search presentation guidance, see the MDN canvas export documentation and Google's guidance on page snippets. Explore the Sycago tools directory for more utilities, or return to the image workspace above to prepare your next batch.

BUILT FOR YOUR NEXT STEP

A lighter image is a good start.
A better website goes further.

Explore Sycago's WordPress website design services for your business.

Discuss your website ↗

Compare your image

Original
Original image preview
Output
Processed image preview

Previews fit the available space. Open a downloaded file at 100% zoom to inspect fine detail.