Select the source files
Pick one image or a batch. PixoSift reads the files in your browser; it does not send them to a PixoSift processing server.
Reduce image file size for websites by choosing an output format and quality setting. Your images are processed in your browser, not uploaded to a PixoSift processing server.
Images are processed locally in your browser and are not uploaded to a PixoSift image-processing server.
Three steps from selecting a file to checking the exported copy. These are visual guides to the workflow, not screenshots of your private files.
Pick one image or a batch. PixoSift reads the files in your browser; it does not send them to a PixoSift processing server.
Use the recommendation as a starting point. Lower quality often makes a smaller lossy file, but may introduce visible artefacts.
Look at text, edges, gradients and fine detail at the size it will appear. Download the result only when it looks good.
No: PixoSift does not overwrite your source file. It creates a separate output for you to download, so you can keep the original unchanged.
However, the exported copy can look different when using lossy compression, and conversion may not preserve metadata or animation. Keep the original and compare the output before replacing an image on your website.
Image compression can make pages lighter and reduce the amount of data a visitor needs to download. The right result depends on the original image, its dimensions, the target format and how much visible change you can accept.
PixoSift lets you process up to 50 images in one batch, adjust the browser encoder quality and compare the original against the output before downloading. It is designed for common website workflows rather than a particular server or framework.
There is no universal percentage reduction: photos, screenshots, graphics and transparent images behave differently. Compare the visual result at the size it will actually appear on your website, and keep the original file until you are satisfied.
Use this guide to choose an initial format and encoder-quality setting. It does not predict a guaranteed file size; always compare the generated image with the original.
For many photographs, WebP quality around 82 is a useful first test. Compare the output before using it.
The quality value is a browser encoder hint, not a universal percentage of visual quality.
Open optimizer with these settingsThere is no format that is best for every image. Use the following as a first decision, then judge the exported file itself.
| Format | Useful when | Watch for |
|---|---|---|
| WebP | General web photos and graphics; can also retain transparency. | Lossy quality can soften fine text or edges; inspect the result. |
| AVIF | You want to test another modern format and your browser can encode it. | Encoding support varies. Confirm that an output file was actually produced. |
| PNG | Lossless output, crisp interface graphics or transparency where exact edges matter. | Files can be much larger than lossy alternatives. The quality slider does not affect PNG output. |
| JPEG | Compatibility with older workflows and photographic images. | JPEG does not preserve transparency. |
A useful first test for photos and many web graphics. It supports transparency and is widely supported by modern browsers.
Trade-off: lossy encoding can soften texture or edges. Compare the exported copy.
Worth testing for photographs when reducing bytes is important; output support depends on the browser's encoder.
Trade-off: if the browser cannot encode it, choose another format. Keep a fallback when publishing to the web.
Useful when lossless output or crisp interface graphics matter. Transparency is supported.
Trade-off: output can be larger; this tool's quality slider does not affect PNG.
A familiar option for photographs and older workflows.
Trade-off: JPEG has no transparency; transparent areas are filled with white by this tool.
Choosing only by percentage: a very small file is not useful if text, edges or gradients look damaged. Leaving oversized dimensions: reducing a 5000-pixel image to a width close to its display size can save more bytes than changing format alone. Overwriting the original: keep the source until you have checked the result in its intended context.
No. Some conversions can produce a larger file, especially when converting to lossless PNG or when the source is already well optimized.
The current tool reads and encodes selected images in your browser. It does not upload them to a PixoSift image-processing server.
Yes. The tool can queue up to 50 images in one batch.
Animated GIFs are not preserved as animation by the current workflow; the browser-decoded frame is converted. Check this before processing animated files.