Images
Resize, convert, compress, crop and inspect images without uploading them.
Written for: Developers and designers · Everyone else, quickly
What this family is for
Image work is usually one of four jobs: making a file smaller, changing its dimensions or format, preparing it for a specific slot (a favicon, a 3×3 grid post, a watermark), or finding out what is actually inside it. Every tool here does that work on your device — a chosen file is read with the browser’s own decoders and re-encoded in the page, never uploaded, so an unreleased photo or a client’s artwork stays on your machine. That also means the tools inherit the browser’s format support: anything it can display, these can re-encode, and a format it cannot decode (HEIC on most desktops, for instance) fails visibly rather than silently.
Resize, convert, compress or embed an image
“Make this image smaller” has three different answers and they are not interchangeable. Resizing changes the pixel dimensions; converting changes the file format; compressing changes how faithfully the pixels are stored at a given size. An image with the same dimensions can drop from 4 MB to 300 KB by being converted to WebP or re-encoded as JPEG at a lower quality, and an image with the right file size can still be the wrong dimensions for where it is going. Decide which property is actually the problem first — the wrong tool fixes the wrong one and loses quality doing it.
The second split is lossy versus lossless. JPEG and WebP are lossy at any quality below 100: re-encoding an already-lossy image loses a little more each time, so convert once from the original rather than in a chain. PNG is lossless, which is right for screenshots, diagrams, logos and anything with transparency, and wrong for photographs, where it produces files several times larger than necessary. Rotating and flipping in 90° steps is lossless for every format; an arbitrary-angle rotation resamples and is not.
The third split is what the output is for. A favicon needs a small square plus a multi-size .ico; a placeholder image is better as SVG text than as pixels; an image embedded in CSS or an email has to become a data URL, which adds about a third to its size in bytes; a photo posted online should have its EXIF metadata checked and usually removed, because it can carry GPS coordinates and a camera serial number. Some of the tools here read or strip that metadata, and some only touch pixels — the page for each one says which.
| If you have… | Use | Because |
|---|---|---|
| Fit an image to an exact width, height or bounding box | Image resizer | Changes pixel dimensions and can convert format in the same step |
| Get under a specific file size in KB or MB | Image compressor | Searches for the best quality that fits, shrinking dimensions only if it must |
| Change JPEG, PNG or WebP without touching dimensions | Image format converter | Format and quality only, with a before/after size comparison |
| Crop to a region or exact pixel coordinates | Image cropper | Draggable selection or numeric crop rectangle |
| Turn an image a quarter turn or mirror it | Rotate & flip | 90° steps are lossless; no resampling |
| Put a label or logo on a photo | Image watermark | Position, opacity, size and tiling, applied locally |
| One image from several | Image stitcher | Vertical, horizontal or grid arrangement |
| Many images from one — a 3×3 grid post | Image grid splitter | Equal tiles with per-tile download |
| A multi-size Windows icon | PNG to ICO | Packs 16/32/48 px into a single .ico |
| A complete favicon set plus HTML | Favicon generator | Every size, an apple-touch-icon, and the markup to paste |
| A stand-in image for a layout | Placeholder image generator | SVG data URL or PNG at any size, with your text |
| An SVG made smaller without changing how it looks | SVG optimizer | Strips comments and metadata and rounds coordinates |
| An SVG as a raster PNG | SVG to PNG | Renders at any size with an optional background |
| An image inlined into HTML, CSS or JSON | Image to base64 | Produces the data: URL, and can read one back |
| To see what the camera recorded | EXIF viewer | Shows the common tags locally; the GPS coordinate block is not decoded into a map |
15 tools in Images
Common mistakes with these tools
- Compressing the same JPEG twice — JPEG is lossy, so each re-encode discards a little more detail. Keep the original and re-export from it; a chain of “just a bit smaller” passes adds visible artefacts.
- Saving a photograph as PNG — PNG is lossless, which is right for screenshots and logos and wasteful for photographs — often several times the size of a high-quality JPEG with no visible gain.
- Resizing to reduce file size when the format is the problem — A 4000×3000 photo displayed at 800 px wide usually needs both. Downscaling alone leaves a JPEG that is still larger than a properly encoded WebP at the same dimensions.
- Assuming a transparent background survives JPEG — JPEG has no alpha channel. Transparency becomes white (or black, depending on the encoder) the moment the file is converted; use PNG or WebP if it matters.
- Posting a photo with GPS metadata intact — Phones embed the capture location, timestamp and sometimes a serial number. Anyone who downloads the file can read it — check the EXIF viewer before publishing, and assume a phone photo carries its location.
- Rotating by an arbitrary angle to “straighten” a scan — That resamples every pixel. Lossless rotation is limited to 90° steps, so a horizon correction costs quality and is best done once, from the highest-resolution source.
Frequently asked questions
What is a base64 data URL?
It is an image encoded as text and written as data:<mime>;base64,<data>. Because it is plain text, it can be pasted straight into an HTML src, a CSS background or a JSON field without a separate file request. The trade-off is size: base64 is about a third larger than the original bytes.
Is my image uploaded?
No. A file you choose is read locally with the browser FileReader and converted in the page; nothing is sent to a server. That is also why the tool works offline once the page has loaded.
Does converting an image lose quality?
It depends on the format you choose. Converting to PNG is lossless — the pixels are re-stored exactly. Converting to JPEG or WebP at a quality below 100 is lossy: the encoder discards detail it judges inessential, and each repeat conversion discards a little more. Always convert from the original file when you can.
Why does my photo get bigger when I compress it?
If the source was already a compressed JPEG, re-encoding it at a high quality can produce a larger file, because the encoder spends bytes re-describing compression artefacts. It usually means the dimensions, not the quality setting, are what need to change — try the resizer or the target-size compressor instead.
Which formats can the tools read?
Whatever the browser can decode — normally JPEG, PNG, WebP, GIF, SVG and AVIF, with format support varying by browser. HEIC from iPhones is not decodable on most desktop browsers, so it fails at the load step rather than producing a wrong result.
Other tool families
- Scripts & numerals — Digits, writing systems and transliteration, converted losslessly. (7 tools)
- Encoding & Unicode — What the bytes and code points of your text actually are. (7 tools)
- Accents & diacritics — Strip the marks that break search, matching and filenames. (1 tools)
- Phonetic & symbol writing — Spell text in other symbolic systems: Morse, Braille, radio alphabet. (3 tools)
- Text utilities — Everyday clean-up: counting, case, whitespace, slugs and lines. (15 tools)
- Developer & data — Encode, format and generate the values developers pass between systems. (24 tools)
- Time & date — Timestamps and dates, converted without a timezone surprise. (4 tools)
- Converters & calculators — Unit, money and grade calculators that show the method, not just the answer. (14 tools)
- Colour — Move a colour between HEX, RGB and HSL without a design tool. (3 tools)
- Documents — Business paperwork — invoices, receipts — generated in the browser. (2 tools)
- CSS & design — Generate gradients, shadows, radii and shapes as copy-ready CSS. (4 tools)
- QR & codes — Machine-readable codes: QR, barcodes, and the subtitle formats players read. (3 tools)