Image to Format Converter

What is an image format converter?

An image format converter takes a picture you already have and re-encodes it as a different file type. A PNG saved out as a JPEG. An AVIF download turned back into something older software will open. A photo re-encoded as WebP so it costs less to send over the wire. The pixels stay the same size. What changes is how they are stored, which decides how big the file is, whether transparency survives, and which programs will accept it at all.

This page converts to four formats: WebP, PNG, JPEG, and AVIF. Three of those are lossy and take a quality setting, defaulting to 90 on a slider that moves in fives. PNG is lossless, so when PNG is selected the slider disappears and the page says outright that quality has no effect. Drop in as many files as you like at once, or click to browse. Each file is handled on its own and comes back as a card carrying the before and after sizes and how much changed, downloadable individually or all at once. The row of format buttons changes from browser to browser. At load the tool checks which of the four yours can encode, and shows only those. If all you need is WebP, the dedicated Image to WebP converter on this site is the simpler page. This one exists for when the format is a decision rather than a given.

When to use it

Nine times in ten it is a file that something else has just refused to accept. You saved an image off a site and got a .webp or .avif. Now the print shop's upload form, the CRM avatar field, an old desktop editor, or a client's document template will only take JPEG or PNG. The image is fine. The encoding is newer than the software trying to read it. Pick JPEG or PNG, drop the file in, download the result, and the upload goes through. The same applies in reverse when a system hands you a batch of BMPs or GIFs that are enormous for what they contain.

The second reason is that you are choosing a format rather than fixing one. A hero image is going onto a page and you want to know whether AVIF or WebP comes out smaller for this particular photo. The answer depends on the image, not on the general reputation of the formats, so encode it both ways at the same quality and read the two numbers. Flat illustrations, screenshots and gradients behave differently from photographs here, and a format that wins on one can lose on the next. Batch drop is useful in the same way: run a folder of product shots through at quality 80, look at the totals, and decide.

Then there are the deliberate one-way conversions. Exporting a transparent PNG to JPEG flattens the transparency onto a solid background. Exactly what you want when a logo has to sit on a white page and the alpha channel is causing more trouble than it solves, and exactly what you do not want if that image still needs to sit over a photo. Converting to PNG is the move when a system has a strict format allowlist, when a downstream tool mangles lossy artefacts, or when you need transparency preserved and the source is a WebP or AVIF with an alpha channel.

How this tool works

Conversion runs through your browser's own image pipeline. Each file is decoded by the browser as an image, drawn onto a canvas at its natural pixel dimensions, and re-encoded using the MIME type of the format you picked. For the lossy formats, the slider value goes along as the quality argument. No library does the compression here. The encoders are the ones already built into the browser you are reading this in, which is why the result for a given format matches what that browser produces anywhere else.

That also explains the changing button row. Browsers do not all ship the same set of canvas encoders, and asking one what it supports is unreliable, so the tool tests instead. It encodes a one-pixel canvas as each format and checks whether the data that comes back really is that format. Formats that fail are dropped from the row, and if the format you had selected was one of them, the selection moves to the first that survived. That probe is doing real work. A browser that fails the test reports no error. It hands back a PNG instead, so without the probe you would be downloading PNGs with an .avif name on the end. WebP, PNG and JPEG pass almost everywhere. AVIF is the one that commonly fails, and Chrome is among the browsers that display AVIF perfectly well while being unable to encode it. Decoding a format and encoding it are separate capabilities.

JPEG has no alpha channel, and this is the trap worth spelling out. When JPEG is the target, the canvas is filled with solid white before your image is drawn on top, so every transparent or semi-transparent pixel is composited against white. A logo with a transparent background comes out on a white rectangle. Soft anti-aliased edges blend toward white rather than staying clean. The white is deliberate, because the canvas default would make those areas black, which is worse. Placed on a white page nobody notices. Placed on a dark background the rectangle is obvious, and there is no undoing it, because the alpha information is gone from the file. Convert to WebP, PNG, or AVIF if transparency has to survive.

Lossy compression is generational. Re-encode a JPEG or WebP that was already compressed and the second round of loss lands on the first, with artefacts compounding even at quality 100. Convert from the highest-quality source you have rather than an export of an export. Metadata does not survive the canvas path. Camera settings, embedded colour profiles, location tags and authorship fields live in the file wrapper rather than the pixels, so none of them cross. A privacy win when stripping location off a phone photo, a real loss when a wide-gamut image comes back flat without its colour profile. PNG ignores the quality slider entirely, because PNG compression is lossless. With no dial to turn, the slider is hidden rather than left doing nothing. Size is the other ceiling: each image is decoded into a canvas in memory, and browsers cap canvas dimensions and area, so a 100-megapixel scan can fail where a 12-megapixel photo is fine. A failed file is logged to the console and skipped rather than stopping the batch.

None of this touches a server. The files never leave your machine. No upload request, no temporary storage, no processing queue somewhere else, just JavaScript reading a file you handed the page and writing a new blob back into your browser's memory. Downloads come from that in-memory blob. For client photography under embargo, internal screenshots, scanned documents, or anything with recoverable metadata, that is the difference between a conversion and a disclosure.

Examples

• Photograph encoded as WebP and as JPEG, both at quality 90

  Input

  1.40 MB PNG

  Output

  437 KB WebP · 419 KB JPEG

  The reason to compare rather than assume. WebP is generally the stronger format, but on this particular photograph at quality 90 the browser's JPEG encoder produced the smaller file by a few percent. Which one wins depends on the image. Grain, gradients, and flat areas all shift the balance. When the file is going somewhere that matters, encode it both ways and read the two numbers instead of trusting a general rule.

• Logo with transparency exported as JPEG, quality 90

  Input

  13.7 KB PNG with alpha

  Output

  17.1 KB JPEG, transparency flattened to white

  The trap case, and note it got bigger rather than smaller. JPEG has no alpha channel, so the canvas is painted white before the image is drawn: the transparent background becomes a solid white rectangle and the anti-aliased edges blend toward white. Fine on a white page, unusable over anything else, and not reversible once saved. The size went up because JPEG is poor at the flat colour and hard edges a logo is made of. This conversion costs you both ways. Do it only when something downstream demands JPEG.

• WebP converted back to PNG for compatibility

  Input

  437 KB WebP

  Output

  3.59 MB PNG

  The file grows eight-fold, and that is what should happen. WebP here is lossy and heavily optimised. PNG is lossless and stores every pixel it is given, including all the compression artefacts baked into the WebP. Re-encoding cannot recover the detail the lossy step discarded. It stores the damaged version faithfully and expensively. Do this only when something downstream will not accept WebP at all; if it will take JPEG, that is a far cheaper way out.

• Screenshot re-encoded as PNG at quality 10 and quality 100

  Input

  67.9 KB PNG

  Output

  122.6 KB PNG at both settings

  Byte-for-byte identical at either end of the slider, which is why the slider is hidden for PNG. The value is never passed to a lossless encoder. The output is still larger than the input even though nothing was lost: the browser's PNG encoder makes different compression choices than whatever wrote the original file, and it optimises less aggressively than a dedicated tool would. PNG round-trips your pixels exactly. It does not promise to match someone else's file size.

Frequently asked questions

• Which is better, AVIF or WebP?

  AVIF usually produces a smaller file than WebP at the same visual quality, particularly on photographs and gradients. WebP is closer on flat graphics and screenshots, and encodes considerably faster. The real deciding factor is reach. WebP is supported almost everywhere, whereas AVIF display support is broad but not universal, and AVIF encoding is missing from a fair number of browsers. Chrome is one of them, which is why the AVIF button may not appear for you here. Where you can encode both, compare your actual image. The winner varies.

• Why did my transparent PNG turn white when I converted it to JPEG?

  Because JPEG has no alpha channel and cannot store transparency at all. To stop transparent areas coming out black, the tool paints the canvas white before drawing your image, so every transparent pixel composites against white and you get a solid white background with edges blended toward white. The alpha is gone from the saved file and cannot be recovered afterwards. Convert to WebP, PNG, or AVIF instead if the image needs to sit over anything other than white.

• Why is AVIF missing from the format buttons?

  Because your browser cannot encode it. The tool tests each format at load by encoding a one-pixel canvas and checking what comes back, then shows only the formats that passed, so the button row reflects your actual browser rather than a hard-coded list. AVIF is the usual absentee. Plenty of browsers, Chrome among them, display AVIF files happily while lacking a canvas encoder for them. Opening this page in a different browser will often produce a different set of buttons.

• Should I use this or the Image to WebP tool?

  The Image to WebP page does one thing, which is convert to WebP, so there is no format to choose and nothing to get wrong. That makes it the simpler, better page when WebP is already the decision. This one adds PNG, JPEG, and AVIF as targets, hides the quality slider when the target is lossless, and probes what your browser can encode. Use the dedicated page for a straight WebP batch. Use this one when you are converting to something else, or comparing formats before you commit.

• Does converting an image between formats lose quality?

  Converting to PNG does not, because PNG is lossless and the pixels round-trip exactly. Converting to JPEG, WebP, or AVIF does, and the loss is generational. If the source was already lossy, this adds a second round of artefacts on top of the first, so always convert from the highest-quality original you have rather than from a previous export. One thing goes in every direction regardless of format: EXIF data and ICC colour profiles are not carried through the canvas, so the output has no camera metadata, no GPS coordinates, and no embedded colour profile.

• Could this run inside our own systems instead?

  Yes, and it is a common request. Converting a handful of images in a browser tab is fine. Converting every asset that lands in a shared drive, normalising product photography on upload, or enforcing one format and quality across a CMS is a job for something that runs on its own, without anyone remembering to do it. Automating that against the systems a team already uses is ordinary work for Zinc Online Solutions. Tell us where the images come from and what has to come out the other end, and we will map out what replacing the manual pass looks like.