Drop a PDF into the zone below and every page comes back as an AVIF image. AVIF lands at roughly half the size of a JPEG at matching quality, and nothing is sent anywhere to get it - your own processor runs an AV1 encoder and the file is assembled here in the page. The tradeoff is time. AV1 does far more analysis per image than JPEG does, so a long document at a high resolution is a real wait. Choose the resolution, convert a page range, take the result as one ZIP or as separate files.
Your files are read by JavaScript running on your own machine. There is no upload, no queue, and no copy on our server to delete later.
Drag a card to reorder it, or use the arrows under each card. The order here is the order in the output. Each card is a preview of the finished page, so what you see is what the PDF will contain.
Esc close ← → move R rotate
Starting
This is your processor doing the work, so the time it takes depends on your machine and not on how many other people are on the site.
The converter above is ready. The guide below explains how a browser produces an AVIF at all - because no browser can, not directly - what an AVIF file is underneath, why an odd sized page comes back one pixel wider than you asked for, and the three limitations worth knowing before you build a workflow on this.
Two minutes, nothing to install, no account for your first run. This PDF to AVIF converter renders and encodes inside the page you are reading, so every step happens on your own hardware.
Click the drop zone above or drag the file anywhere onto this page. The document is parsed as it lands and the card shows page one with the page count. If your browser cannot encode AV1 the tool says so before it renders anything, rather than after.
A PDF page has no resolution of its own, so this is your decision. 150 dpi suits most screen work. Encoding an AV1 frame is genuinely expensive in processor time, so start with a modest resolution and a short page range to find out how fast your machine handles it before you commit to a whole document.
Quality 85 is the recommended default. AVIF holds together at lower settings far better than JPEG does, so 75 is often still presentable on a page of type. Use the Pages field to convert only what you need.
Press Convert to AVIF. Expect it to take noticeably longer than JPG or WebP, because AV1 is a video codec doing a great deal more analysis per frame. Pages are encoded one at a time and the progress readout names the page it is working on.
One fact shapes this entire tool. No browser can encode AVIF. Ask a canvas for an AVIF and it hands you back a PNG, quietly, with the wrong MIME type attached and no error of any kind. Measured in current Chrome, the browser's own image encoder supports PNG, JPEG and WebP, and nothing else.
That is worth knowing because it means any tool claiming client-side AVIF export is either doing something specific and deliberate, or it is handing you PNG data in a file named .avif. There is no third option.
This tool takes the long route, and it goes like this.
What comes out is a genuine AVIF that any AVIF reader will open, produced without a single byte leaving your machine and without downloading a multi-megabyte encoder library first.
Why the color space is read and not assumed. The range flag alone shifts every pixel by sixteen levels. Writing a guess produces a file that opens perfectly and looks washed out or crushed, which is worse than a file that fails outright, because nobody ever investigates a file that opens.
Decoding AVIF is universal now. Encoding it in a browser requires WebCodecs with an AV1 encoder, which in practice means Chrome, Edge, Opera and the other Chromium-based browsers. The check runs when you press convert and the tool refuses with a plain explanation rather than producing a broken file. On a browser without it, PDF to WebP gets you most of the size benefit and works everywhere.
AV1 buys its compression by trying far more possibilities than JPEG does: many block partitionings, dozens of intra prediction modes, several transform types, each one evaluated and scored. That search is where the small files come from. Expect a page to take several times longer than the same page as JPG, and expect a long document at 300 dpi to be a genuine wait. Start with a short page range and calibrate against your own machine.
AV1 encodes color at half resolution in each direction, a scheme called 4:2:0 chroma subsampling. There is no way to represent half a chroma sample, so odd pixel dimensions have no clean mapping and encoders pad them unpredictably. Rather than gamble on that, this tool draws the page onto an even sized canvas, fills the extra row or column with white, and declares the even size in the container.
So a page rendering to 595 by 842 comes back as 596 by 842. The extra pixel is white and sits at the edge of a page margin, where nobody will ever see it. It is stated here because a workflow that asserts on exact dimensions needs to know, and because a converter that changed your dimensions without mentioning it would be doing something worse.
AVIF is the most efficient still image format in general use. It routinely produces files around half the size of a comparable JPEG, meaningfully smaller than WebP, and it degrades into softness rather than falling apart into visible blocks. That efficiency earns its keep in some places and is irrelevant in others.
Serve it with a fallback. An AVIF source inside a <picture> element with a WebP or JPEG alternative gives every visitor the best format their browser reads, and it costs nothing to implement. Publishing AVIF as the only format is the one deployment mistake here worth avoiding.
This is unusual for AVIF in particular. Because browsers cannot encode the format, virtually every AVIF conversion service in existence does the work on a server, which means uploading your document to somebody else's machine and trusting a policy document about what happens to it once it arrives.
There is no upload step here. The page is rendered and the AV1 frame is encoded by your own processor through WebCodecs, and the container is assembled in JavaScript on this page. No server-side component takes part, so no copy of your document exists anywhere to be breached, retained past a stated window, subpoenaed, or found in an unsecured bucket in three years.
Verify it rather than believe it. Load this page, disconnect from the internet completely, and convert. It works, because there is nothing left for it to talk to.
Thirty seconds to prove it. Open this page, turn off your Wi-Fi, drop in a PDF and convert. Every server-based AVIF converter fails at that point. This one produces the images.
The only thing that reaches a server is the operation counter enforcing the free tier. It records that a conversion happened and when. It never receives your document, its name, its page count, or anything derived from its contents.
By taking a longer route. The page is rendered to a canvas, that canvas is encoded as a single AV1 intra frame using WebCodecs, which is the low level media interface built into the browser and does have an AV1 encoder, and the AVIF container is then assembled here box by box. The codec configuration record that the container requires is parsed directly out of the sequence header the encoder produced, because WebCodecs does not supply one for AV1.
Because AV1 buys its compression by searching. It evaluates many block partitionings, dozens of intra prediction modes and several transform types, scoring each one, and that search is what produces a file half the size of a JPEG. JPEG does almost none of it. The extra time is where the saving comes from, which is why you should start with a short page range and see what your machine does.
Client-side AVIF encoding needs WebCodecs with an AV1 encoder, which today means Chrome, Edge, Opera and the other Chromium-based browsers. The check runs before any work starts, so you are told rather than handed a broken file. On a browser without it, PDF to WebP gives you most of the size benefit and works everywhere, and PDF to JPG works on absolutely anything.
AV1 encodes color at half resolution in each direction, and there is no such thing as half a chroma sample, so odd pixel dimensions have no clean mapping. Rather than let the encoder pad them unpredictably, the page is drawn onto an even sized canvas with the extra row or column filled white, and the container declares that even size. The extra pixel sits in the page margin and is invisible, but it is stated here because a workflow asserting on exact dimensions should know.
Roughly half at matching visual quality, and meaningfully smaller than WebP as well. It also degrades far more gracefully. Where JPEG falls apart into visible blocks at low quality, AVIF goes soft instead, which means quality 75 is often still presentable and the saving widens further.
No, and for AVIF that is genuinely unusual. Because browsers cannot encode the format, nearly every other AVIF converter runs on a server and requires you to upload your document first. Here the rendering, the AV1 encoding and the container assembly all happen on your machine. Confirm it by disconnecting from the internet and converting anyway.
In a browser, yes. Outside a browser, support is still patchy: plenty of office suites, image viewers, print pipelines and document systems do not read it. Use AVIF for web delivery, and serve it inside a picture element with a WebP or JPEG fallback. For a file somebody will open on their own computer, send JPG or PNG.
No. The output is lossy AV1 at the quality you select, which is what the format is used for and where its efficiency comes from. If you need a lossless page render, use PDF to PNG, which discards nothing and on flat document pages is often surprisingly competitive on size.
Yes, and with AVIF it is worth doing. Type a range into the Pages field the way you would in a print dialog: 1-5, or 2,7,11-14. Because the encoding is slow, converting the four pages you need rather than the eighty you have is the difference between a few seconds and a long wait.
For web delivery, start from your layout: decide how wide the image will display, double it so it stays sharp on a high-density screen, and pick the dpi that reaches that. 96 dpi gives an A4 page at 794 pixels and covers most preview slots. Higher resolutions multiply the encoding time as well as the file size, and past 900 million output pixels across a run the tool refuses the job and asks you to lower the dpi.
Several pages come back as a single ZIP by default, each named after the source document with a zero-padded page number so they sort correctly. Switch Deliver as to Separate files for individual downloads. A single page always arrives as a plain image rather than an archive.
No. Nothing is stamped on the output, no logo is added, no trial banner appears and no attribution is required. The images contain your pages and nothing else, and they are free for commercial use.
On a Chromium-based mobile browser it can, but it is the least comfortable of the six formats there. Encoding is processor intensive, phones throttle hard under sustained load, and a phone tab has far less memory headroom than a desktop one. Keep the resolution modest and the page range short, or use PDF to WebP, which is much lighter work for a similar result.
Not for your first operation. One conversion run counts as one operation regardless of how many pages it produced. Anonymous visitors get one per 12 hours counted by network address, a free account gets two, and Premium removes the counter and raises the batch ceiling from 20 files or 100 MB to 100 files or 500 MB.
All of these run the same way: your file is read by JavaScript on your own machine, the work happens on your processor, and the result is handed back to you as a download that never existed anywhere else.
Find the apps that slow your store down, by name and in milliseconds, then answer shoppers' doubts with trust badges under Add to cart. Both are free, with no card and no access to your store admin.
Paste any page and get its three Core Web Vitals, the apps costing time, like Klaviyo or Judge.me, and the Shopify or WooCommerce fix.
Show the promises that settle doubts, from free shipping to a money-back guarantee, right under the Add to cart button. One line in your theme.
Change the link after printing and count the scans. 5 tracked codes on a free account, static codes unlimited.
Anonymous visitors get one operation per 12 hours, per network address. That is not a sales tactic, it is the only lever we have: this tool runs on your processor, not ours, so the run itself earns us nothing. An account tells us who you are, and the ceiling goes up the moment you have one. Your allowance resets in 12 h.
No credit card, no trial period, no sales call. Join with Google, GitHub, Facebook or email.
Free accounts get two operations per 12 hours, and yours reset in 12 h. Waiting costs you nothing and is a completely reasonable choice. If PDF work is a weekly job rather than a one-off, the counter is the wrong tool for you and Premium takes it off the page.
Cancel anytime. Everything you have already produced stays yours either way.