Quick Image Saving: Bulk Image Downloader with CSS Selector

Save page images in bulk: scan and pick visually, or match exactly with a CSS selector. Originals, ZIP, all local.

Quick Image Saving takes the images off a web page and saves them to your computer as a ZIP file, and photographs the page itself. It opens in Chrome's Side Panel, so it stays put while the browser's save dialog is on screen, and a download keeps running even if you close the panel. About the permissions, before you install Chrome lists three: access the page debugger backend, read and change all your data on all websites, and manage your downloads. Two features account for all three, and it is worth knowing which. The first two are the screenshot tool. Chrome's debugger interface is what lets the extension preview a page at another width and photograph a whole page in a single pass - the same interface the browser's own device mode uses - and Chrome attaches its broadest data warning to that interface whether or not an extension also asks for website access. This one does not: its manifest requests no website access at install. Chrome does not allow the debugger permission to be requested later, on first use, so it has to be part of the install. The third is exactly what it says - saving your ZIP and your screenshots into your downloads folder. While a screenshot session is open, Chrome shows its own bar above the page saying that Quick Image Saving started debugging this browser. That bar is Chrome's, not ours, and it is expected. It appears when you press a screenshot button, stays while you set the width, and goes as soon as the picture is saved or you press Cancel. Nothing is read from the page through it beyond the screenshot you asked for, and the image goes straight to your downloads folder. Website access is granted separately, and none of it at install. When a download needs to read images from a host, Chrome asks you about that host at that moment, and only about the hosts of the images you picked. You can see and narrow what it holds at any time in chrome://extensions, under Site access. Two ways to save images Simple Mode - press Scan this page. Everything found is listed in a grid with a thumbnail, its size in pixels, its format and where it came from. Filter by size, format or source, search the addresses, tick what you want, and save it as a ZIP. Advanced Mode - type a CSS selector, such as article img, #gallery img, or img[src*="/uploads/"], and save exactly what it matches. Recent selectors are remembered; delete them one at a time or clear them all. Pick an area on page - outline one gallery or section with the mouse or with the arrow keys, and scan only that. What a scan finds - srcset and picture renditions, not only the one currently on screen - lazy-loading containers that keep the real address in a data attribute - CSS background images, including layered backgrounds and image-set() - links that point straight at an image file, which on a gallery is usually the full-size version behind the thumbnail - images inside open shadow roots Originals, not the display copy Many sites keep a photo as PNG or JPEG and hand your browser a smaller WebP or AVIF copy of it. Quick Image Saving reads the addresses the page publishes and puts the original format ahead of the converted one, so the file in your ZIP is the original. There is nothing to switch on. Inside one format it takes the largest rendition the page offers rather than the small one that happened to be displayed. When a site publishes nothing but WebP, that WebP is saved, so a download never fails for this reason. Some hosts convert an image and keep the old .png or .jpg address, so the address alone cannot be trusted. Every saved file is named after the bytes that actually arrived, and the report inside the ZIP points out which files those were. Every ZIP explains itself Each archive holds a short text file, quick-image-saving-report.txt. It records the page the images came from, the selector and rules used, how many files of each format are inside, and the web address of every single file - including the ones that had to fall back to another address and the ones that needed a retry. Weeks later you can still tell where any image came from. Downloading - Four images are fetched at a time, and a request that failed in a way that looks temporary - a timeout, a rate limit, a server error, a dropped connection - is tried again, at most twice. A 404 or a 403 is not retried: the first answer already said everything. - Large jobs are split into several ZIP files, by number of images and by their size - Before a large download starts you are told how many images and roughly how many megabytes it will be, and asked to confirm - Keep original names is optional; duplicate names get -2, -3 and so on - Convert format re-encodes saved images to PNG or JPG on your own machine, from the original the extension downloaded - a file already in that format is left exactly as it arrived, and the site is never asked for a different format. JPG puts transparent areas on white. - URL rules rewrite image addresses before the download, for example a thumbnail address into the full-size one. A rule is scoped to a site, rules run in the order you set, there is a live preview, and Test transformed URL really fetches the result before you save the rule. Simple Mode uses those same rules through one switch, Prefer original quality, with no regular expressions to write. Page screenshots - A fixed header appears once instead of on every screen, and the finished image has no visible seams - Or only the visible part, or an area you outline on the page - A switch closes up the long flat stretches a pinned, scroll-driven section leaves behind, and the panel says how many pixels it took - Set the page width anywhere from 320 to 3840 pixels, with Mobile, Tablet, Laptop and Desktop presets, a slider, or an exact number. Only the page is emulated - your window and the panel stay where they are. - Zoom between 20% and 400% to check the layout, and once you are zoomed past the point where the whole width fits, drag the page to move around. The saved image ignores the zoom. - Hide floating elements keeps fixed bars and cookie banners out of the shot; the page gets them back immediately afterwards - Long pages are walked through before the shot, so lazy-loaded images are in the picture instead of grey boxes - Pages whose content only draws while it is on screen are photographed a screen at a time, and a switch asks for that outright. A full-page shot is also read back once it has been put together: photographs that came back as the bare background behind them are counted from the picture itself, and where the page can be scrolled the shot is taken again a screen at a time. The panel says it retook the page screen by screen, or says how many photographs came back empty, or both - The file is a PNG named after the site, the date, the kind of shot and its pixel size. A page too long to fit in one image is saved down to that limit, and the panel tells you where it stopped. What is new in 1.0.0 The Chrome Web Store has no release-notes field, so it is here. - The interface moved from the toolbar popup into Chrome's Side Panel, where it survives the save dialog and can be closed without stopping a download - Simple Mode: scan a page, review the images in a grid, and download without writing a selector - Page screenshots: the whole page, the visible part or an area you pick, at any width from mobile to 4K - Pick an area on the page instead of describing it with a selector - Scanning now finds CSS backgrounds, linked originals, lazy containers and open shadow roots - URL rules with a preview and a real fetch test, and Prefer original quality in Simple Mode - Convert saved images to PNG or JPG locally, leaving alone anything already in that format - Downloads run four at a time with retries, and archives split by size as well as by count - Counts of your own use, kept on your device and shown to you in Advanced Mode - A light theme beside the dark one Updating from 0.0.0.9 This version needs the debugger permission for the screenshot tool, and Chrome does not let an extension ask for that one later, so it is declared at install. Because that permission carries Chrome's own warnings, Chrome asks you to approve them before the updated extension runs. Nothing stored is lost when you do: your selector and regex history, your saved rules, your options and your light or dark theme are all still there. Website access is worth a word, because the warning you see does not change. Chrome shows "read and change all your data on all websites" for both versions, but for different reasons: 0.0.0.9 genuinely required access to every site, while 1.0.0's manifest requests none and that line comes from the debugger permission the screenshot tool needs. If you are updating, the site access you already granted stays granted - Chrome carries it over - and you can narrow it whenever you like at chrome://extensions, under Site access. Privacy Everything happens on your machine. Images are matched, fetched, archived and converted in your browser, and screenshots are rendered and saved there too. Image requests go straight to the websites that host those images, exactly as they would if you saved each picture by hand; nothing is proxied through Dev Raiks. There is no account and nothing to sign in to. There is no analytics of any kind: your selectors, rules, history and local counts live in Chrome's extension storage on this device and are sent nowhere. Good to know - Works on ordinary http and https pages, and only those. Clicking the toolbar icon on a chrome:// page or on a local file connects nothing, so scanning and screenshots are unavailable there; the panel will keep asking you to reconnect the page, because from its side the page was never connected. Chrome blocks extensions on its own Web Store pages as well, so nothing runs there either. - If nothing downloads, the page itself may be blocking the requests. Cross-Origin Mode, in Advanced Mode, asks Chrome for access to the image hosts your selection needs and fetches them from the extension instead. - The extension does not get around logins, paywalls, DRM or a site's access rules, and it does not change what you may do with an image. Please respect copyrights, licences, site terms and the law. Developer: Dev Raiks - https://devraiks.com/ Privacy policy: https://devraiks.com/quick-image-save/privacy-policy/