How to Convert Images: PDF, SVG, Vector, Text

How to convert images depends entirely on what you mean by convert, because it is never one job. Turning a photo into a PDF wraps it, turning it into an SVG traces it, and pulling the words out of a scan reads it with OCR. Pick the file you have, then the route follows.

That is the trap in this whole topic. People search convert image to PDF and convert an image to a vector as if they are the same task, land on a tool that quietly does the wrong thing, and walk away with a PDF that will not search or a JPEG wearing an .svg extension. Each conversion has a real mechanism behind it, and once you know which one you need, the right tool is usually free and already on your machine. This page is the map, and the eleven guides linked below carry each walkthrough.

How to convert images: start with the file you have

The reliable way to convert images is to name the operation first, because convert hides at least three different jobs that need three different tools. Changing a JPEG into a PDF wraps it in a page. Changing it into an SVG traces it into drawing commands. Pulling the words out of a scan reads it with OCR. Get the operation right and the tool is almost always free and already installed.

Here is the split that decides everything else on this page.

A raster image, also called a bitmap, is a fixed grid of pixels: a JPEG, PNG, GIF, WebP, or AVIF (MDN image type guide). A vector image stores drawing commands instead of pixels, which is why an SVG scales to any size, from a favicon to a billboard, without turning blurry or blocky. A PDF is neither, it is a container, a page wrapped around your picture that changes nothing about the image inside. Once you know which of those three you are producing, skip to the matching section below.

What does converting an image actually mean?

Converting means changing the container, the encoding, or the underlying structure of a file, and those are three separate problems that people wrongly treat as one. Naming which one you are doing is what stops you from reaching for a tool that produces the wrong result.

Raster and vector are not interchangeable

A photo is raster: millions of pixels, each holding one color, with no notion of shapes or lines anywhere in the file. You cannot losslessly turn it into a vector, because there are no drawing commands hidden in those pixels to recover. The most you can do is trace an approximation, which is a different operation from a format swap. SVG images are defined in XML text, so they can be searched, indexed, scripted, and edited in any text editor, and they render at any size without loss of quality (MDN SVG). That is exactly what a fixed grid of pixels cannot do.

A container is not an image format

Putting a JPEG into a PDF does not change the image at all. An image-to-PDF tool preserves the original quality and dimensions of whatever you feed it; the PDF is a page wrapped around your picture, nothing more. This is why a PDF built from a scanned page is not automatically searchable: unless something has run OCR on it, the file is still a picture of text, not text you can select. Keep that distinction in mind and most conversion confusion disappears.

How do you convert an image to PDF?

Use the print function already built into your computer before you install or upload anything, because every desktop operating system can print an image straight to a PDF for free. The online route is worth it only when you need to merge several images or you are working on a phone.

On a Mac, print then save as PDF

Open the image, choose File, then Print, click the PDF pop-up menu at the bottom left of the print dialog, and choose Save as PDF; give it a name and a location, and add a password under Security Options if the document is sensitive (Apple Support). There is no download and no account, and it works with any image Preview can open, which is effectively all of them.

On Windows, use Microsoft Print to PDF

Open the image in Photos or any app that can print, choose Print, and pick Microsoft Print to PDF as the printer instead of a physical one. It ships with Windows and writes a PDF file wherever you point it. The exact button labels have shifted between Windows versions, so the dedicated walkthrough keeps the current clicks rather than a snapshot that goes stale.

On a phone, or combining several images

For a phone, or to merge a stack of pictures into one document, a browser tool is usually faster than hunting through a print menu. Adobe's online tool accepts JPG, PNG, BMP, GIF, and TIFF, and it will combine images into one PDF automatically when you upload several, even mixed formats; the free tier caps uploads at 100 MB and discards the file unless you sign in and save it (Adobe image to PDF). If the source was a scanned page and you need to search the text inside it, that is a separate OCR step, covered further down, not something a plain image-to-PDF conversion does for you.

Shrink the source before you wrap it

One habit saves the most trouble here: compress the image before it goes into the PDF. A PDF built from a 4 MB photo is a 4 MB PDF with a cover page, because the wrapper adds almost nothing on its own. Run the picture through a real compressor, the way the compress-images guide walks through, and the PDF comes out proportionally smaller with no visible change to the image. We do not build or edit PDFs, and this page says so plainly rather than pretending; our tool only handles the image going into one, making it smaller and cleaner first. It is also worth checking the page size and orientation the PDF tool picks, since a tall phone photo dropped onto a wide default page leaves large empty margins that add nothing but confusion.

How do you convert an image to SVG or vector?

You trace it: software scans your pixels and generates paths that approximate the shapes, because there is no way to reconstruct real drawing commands from pixels that never had any. That is the whole reason a photo cannot become a true vector.

Tracing with Inkscape, which is free

Import the image, select it, and open Path, then Trace Bitmap, with the shortcut Shift+Alt+B. Choose a Single scan for a one-color result, which suits logos and line art, or Multiple scans to build layered paths for artwork with more color, then preview and apply. The engine underneath is Potrace, and Inkscape can run SIOX first to separate a subject from its background before it traces (Inkscape manual, tracing bitmaps). Adobe Illustrator's Image Trace does the same job inside the paid suite if you already own it. A clean traced logo is often only a few kilobytes of text, and because it is defined as paths rather than pixels it stays razor sharp on a phone, a retina laptop, and a printed banner from the one file, which is the entire point of going vector in the first place.

When tracing works, and when it does not

Tracing is good on flat, high-contrast art: logos, icons, line drawings, anything with clean edges and a small number of colors. It is bad on photographs, and that is the single most common misfire in this whole category. You can embed a photo inside an SVG file, but that only wraps a bitmap in XML; it does not vectorize anything, the same non-conversion trick as stuffing a JPEG into a PDF. If your source is a photo and you need it to scale forever, vector is the wrong format no matter which tool you throw at it. Deciding to convert an image to a vector is really a decision about the source art, not just a file operation, so judge the art first.

How do you convert an image to text with OCR?

Optical character recognition, OCR, reads the shapes of characters in a picture and turns them into machine-readable text you can select, copy, search, and edit (Adobe OCR), which a plain scan or screenshot can never do on its own no matter how you open it.

The quickest test for whether you need OCR is to open the file and try to select a sentence with your cursor. If nothing highlights, you are looking at a picture of text, and it needs OCR before it is useful for anything beyond reading on screen. The online path is the same shape for most services: upload the scan, OCR runs automatically, and you download a searchable version, usually after making a free account. Desktop tools such as Acrobat Pro can OCR straight from image files like PNG, JPG, and TIFF, not only from existing PDFs. Whichever tool you use, read the output back before you trust it: most engines let you set the document language, and picking the right one lifts accuracy noticeably on anything that is not plain English, while numbers, tables, and stamps are where errors hide.

Accuracy is not a guarantee, and this is where people get burned. Low resolution, unusual fonts, and handwriting all degrade the result, and a higher-resolution scan measurably improves what comes out the other side. If a document matters, scan it well the first time, because there is no OCR setting that recovers detail a blurry capture never recorded. When you convert an image to text, you are betting on the quality of the input, so put your effort there.

How do you convert between JPEG, PNG, WebP, and AVIF?

This is the one conversion this site's own tool actually does: drop in a JPEG, PNG, WebP, or AVIF and pick a different one of those four as the output format, with the compression handled in the same step. The reason it matters is that the four formats are not interchangeable.

JPEG is lossy and built for photographs. PNG is lossless and right for screenshots or anything needing transparency. WebP typically lands 25 to 35 percent smaller than a JPEG at the same visual quality, in both its lossy and lossless modes. AVIF usually goes further, often around 50 percent smaller than JPEG, at the cost of slower encoding. Here is how the sizes stack up on the same photo.

The measured example on this site makes the range concrete: the homepage demo file, a 497 KB JPEG, becomes a 214 KB WebP at quality 80, a 56.9 percent cut with no visible loss at normal zoom.

Converting does not undo damage

Going the other direction will not restore anything. Saving a JPEG as a PNG does not bring back detail the JPEG already discarded; you get a larger, lossless file of the exact same degraded image. If you need the format changed, convert and compress between them here in one step; if you need the quality back, you needed a better original, not a different extension. That asymmetry, lossy loss being permanent, is the rule that catches the most people.

Transparency is the other trap in this group. JPEG has no transparency channel, so converting a PNG or WebP logo with a see-through background to JPEG fills that background with a solid color, usually white, and the change is silent until you place the image on a colored page. If the transparency matters, convert to PNG or keep WebP; if it does not, JPEG is fine and much smaller.

What are the most common conversion mistakes?

The mistakes in this category almost all come from treating convert as a single magic verb instead of naming the real operation. Five come up again and again.

A PDF is not an image format, it is a container, so convert to PDF never changes the picture inside, only what wraps it; if you needed a smaller file, you needed to compress the image before it went in. Renaming or embedding a JPEG as .svg does not vectorize it either; without an actual trace step you have a bitmap wearing a vector extension, and it will pixelate at large sizes exactly like the JPEG did. SVG is the wrong tool for photographs specifically, not a universal upgrade over raster formats, so a traced photo comes out rough and strange. If you cannot select the text in a PDF, assume it is a scanned image until proven otherwise, because that is almost always what it is, and it needs OCR. And converting a JPEG to PNG for better quality gets you a bigger file with the same already-lost detail, since lossless compression cannot recover what lossy compression threw away upstream. A sixth mistake is quieter: assuming an online converter keeps your file private. Many upload your image to a server and hold it for hours, so for anything sensitive prefer the offline OS methods above, or a tool that states plainly how long it keeps files and deletes them on a short clock.

Which conversion do you actually need?

Match what you have to what you actually want, not to the word convert, and the right route and the right guide fall out immediately.

Two guides pair well with this hub. JPEG explained covers converting to and from the format everything else gets compared against, and WebP and AVIF covers the two formats most likely to be the right answer when file size matters more than compatibility with an old viewer.