AVIF is a modern image format built on the AV1 video codec. At the same visible quality it typically produces files much smaller than JPG, which is why it has become a favorite for speeding up web pages. Converting a JPG to AVIF is a common step when you want your existing photos to load faster without looking any worse.
There is one honest caveat worth stating up front. Your JPG is already a lossy file, meaning some detail was thrown away when it was first saved. Re-encoding it as AVIF is a lossy-to-lossy conversion, so you are compressing already-compressed data. The good news is AVIF is efficient enough that a sensible quality setting keeps this invisible while still shrinking the file a lot.
What you gain, and the one catch
The gain is size. For a typical photo, an AVIF can be a fraction of the JPG's weight at a quality most people cannot tell apart. On a page full of images, that adds up to faster loads and less data used, which is the entire point on the modern web.
The catch is that you are not recovering anything the JPG lost. AVIF cannot put back detail that was already discarded; it can only preserve what is left more efficiently. That is fine for the goal here, which is a smaller file that looks the same, not a higher-quality one. Just do not expect the AVIF to look better than the JPG you started from.
AVIF makes your JPG dramatically smaller at the same look. It cannot improve quality, only preserve it more efficiently.
JPG and AVIF: two generations of compression
Both formats are lossy, but they come from very different eras of image coding, and that gap is the whole reason the conversion is worthwhile. JPEG was standardized around 1992 as ISO/IEC 10918, and it compresses by splitting the image into eight-by-eight pixel blocks and applying a discrete cosine transform to each one, discarding the high-frequency detail your eye is least sensitive to.[1] It was a brilliant design for its time and it still works, but its block-based approach shows its age: push it hard and you see the classic blocky artifacts and ringing around sharp edges. In everyday files the JPEG data is wrapped in the JFIF interchange structure that gives the format its familiar on-disk layout.[2]
AVIF is the image counterpart of the AV1 video codec, defined by the Alliance for Open Media, which packages a single AV1-coded keyframe inside the same ISO base media container that modern video uses.[3] AV1 brings decades of newer coding technique to the table: much larger and variable block sizes, far more sophisticated intra-frame prediction, and a modern entropy coder. The practical upshot is that AVIF represents the same visible image with dramatically less data than JPEG, and it degrades more gracefully when you do push the quality down, softening rather than shattering into blocks.[4]
AVIF also brings capabilities JPEG never had: it can store full alpha transparency, it supports higher bit depths for smoother gradients, and it handles wide color gamuts and high dynamic range. None of those help when your source is an ordinary 8-bit JPG, but they explain why AVIF is treated as a genuinely next-generation format rather than just a smaller JPEG.
The efficiency gap is not just theory. JPEG's discrete cosine transform operates on fixed eight-by-eight blocks and has essentially one tool for predicting content, whereas AV1 can choose from many block sizes and dozens of intra-prediction modes, picking whatever describes a given region most compactly. Because a modern encoder spends real computation searching for the best combination, an AVIF encode is far slower than a JPEG encode, sometimes by orders of magnitude. That cost is paid once, by you, at conversion time; the benefit, a smaller download, is collected by every visitor thereafter. On the web that maths almost always favors AVIF, which is the core reason the format has spread so quickly despite being harder to encode.
How the re-encode works under the hood
A JPG-to-AVIF conversion is a full decode-and-re-encode, not a repackaging. First the tool decodes the JPG: it reverses the JPEG process, running an inverse discrete cosine transform on every block to rebuild a grid of raw pixels in memory. At this moment you have the picture exactly as the JPG could represent it, artifacts and all, with no further loss than the original file already carried.
Then the tool re-encodes those pixels with the AV1 image encoder at the quality you asked for. The encoder analyzes the image afresh, chooses its own block partitioning and prediction, and quantizes the result to hit your target. Because this is a second lossy pass over data that was already lossy, it is technically a generation-loss step. The reason it rarely matters in practice is that AVIF is efficient enough to reproduce the JPG's appearance, including its existing artifacts, at a quality setting that still yields a much smaller file.
Generation loss and why it stays invisible
Whenever you save a lossy image as another lossy format, you add a second round of approximation on top of the first. In theory that stacks: each generation can smear detail a little more. AVIF keeps this invisible for two reasons. First, its coding is precise enough that at a sensible quality the encoder can track the pixels it was handed very closely, so the second approximation is far smaller than the first one the JPG already made. Second, the artifacts that a JPG carries are smooth, low-detail regions to an AVIF encoder, and smooth regions are exactly what these codecs compress most faithfully.
The one situation where generation loss can bite is if you set the AVIF quality too low. Then the encoder starts discarding real detail on top of the JPG's existing losses, and softness or smearing appears. The defense is simple: pick a moderate quality, and compare the AVIF against the JPG at full zoom before committing, which is the habit described below.
Convert JPG to AVIF, step by step
Open the converter and add your JPG
Open the FileFormer image converter and drop your JPG in. The re-encoding runs on your device, so nothing is uploaded.
Choose AVIF as the output
Set the target to AVIF. This is the modern, high-efficiency format that will shrink the file the most.
Set the quality
A quality around 50 to 65 usually looks identical to the source while cutting size hard. AVIF holds up better at lower numbers than JPG does, so you can push it further.
Convert and download
Convert, compare it to the original at full zoom, and download. If you spot any softness, nudge the quality up and export again.
Choosing a quality that stays sharp
Because the source is already lossy, the safest habit is to check the AVIF at full size against the JPG before you commit. Photos with smooth areas like skies compress especially well. Images with fine, high-contrast detail are where over-compression shows first, so if your JPG is text-heavy or full of sharp edges, keep the quality on the higher side.
Serve AVIF with a JPG or WebP fallback on your site. Support is strong in current browsers but a fallback keeps older ones covered.
Settings that shape the result
Beyond the single quality number, a few controls change how an AVIF turns out. Knowing what each one does helps you tune the trade between size, sharpness, and how long the encode takes.
| Setting | What it controls | Practical guidance |
|---|---|---|
| Quality | How much detail is kept versus discarded | 50 to 65 is usually indistinguishable; raise it for text and hard edges |
| Encoder effort or speed | How hard the encoder searches for the best coding | Slower effort gives smaller files for the same quality; worth it for images you save once and serve many times |
| Chroma subsampling | How much color resolution is kept relative to brightness | 4:2:0 saves size on photos; keep 4:4:4 for crisp colored text or fine graphics |
| Bit depth | Precision of each color channel | 8-bit matches a JPG source; higher depth only helps if the source has it |
The one to actually spend time on is encoder effort. Because you typically encode a web image once and then serve it to thousands of visitors, letting the encoder work harder for a smaller file is almost always a good trade. The size win carries every single time the image is downloaded.
Real-world scenarios and the right approach
The best quality setting and the decision to convert at all depend on what the JPG contains and where it will be used. A few common cases make the trade concrete.
| Scenario | Recommended approach |
|---|---|
| Hero or banner photo on a landing page | Convert at quality 55 to 65 with slow effort; the size win multiplies across every visitor |
| Product photos in a catalogue | Batch convert at a fixed high-ish quality; keep 4:4:4 if any have fine text or logos |
| A blog full of inline JPGs | Convert but serve with a JPG fallback so older clients still see images |
| A photo emailed to a client | Skip AVIF; many mail clients cannot render it, so keep the JPG |
| A screenshot with sharp UI text | Use high quality and 4:4:4, or consider a lossless format instead of AVIF |
The recurring lesson is that AVIF pays off most on photographic images you control and serve repeatedly, and pays off least when the file has to travel into someone else's software. The closer a case sits to "my web page, my images, many visitors," the stronger the argument for converting; the closer it sits to "handing a file to an unknown recipient," the more a plain JPG remains the safe choice.
When AVIF is not the right target
Skip AVIF if the image has to open in old software or be embedded somewhere that only accepts classic formats; those places still expect JPG. If you want strong compression with the widest possible support today, converting JPG to WebP is a gentler step that almost everything reads. Reach for AVIF when you control the page and want the smallest modern files you can get.
It is also worth being clear about what AVIF is not for. It is a delivery format, not an editing master: because the conversion is lossy, you should always convert from your best available JPG rather than re-editing an AVIF and re-exporting it, which would stack yet another lossy generation on top. And AVIF gives you nothing on an image that is already tiny, such as a small icon, where the file is dominated by fixed overhead rather than pixel data. The sweet spot is a reasonably large photograph that will be downloaded many times, which is exactly the case the modern web is full of, and exactly why converting existing JPGs to AVIF has become such a common performance win.
Convert your JPG to AVIF now
Pick AVIF, set the quality, and download, all in your browser with nothing uploaded.
Key takeaways
- AVIF makes a JPG much smaller at the same visible quality.
- It is a lossy-to-lossy conversion, so it preserves quality rather than improving it.
- A quality around 50 to 65 is usually a clean, small result.
- Use WebP or a fallback where the widest compatibility matters.