Ruxget privacy notice

Applies to: Chrome extension 0.4.81 and companion desktop app 0.4.81. Last updated: 2026-11-09.

Ruxget detects media on web pages, downloads video or audio, and merges MP4 or encodes MP3 inside Chrome. Ruxget has no account system of its own, no ads, no analytics SDK, and no developer server that receives user data.

The floating button added in 0.4.2 only reads the player’s position and shows itself on Bilibili video pages where Ruxget already has permission. Clicking it opens the Ruxget panel; it does not start a download or add website access. The button is hidden while detection is paused, and you can also hide or restore it with a switch in the extension. Desktop batch tasks, category settings and restore backups stay on your computer and are not uploaded.

The desktop details view added in 0.4.4 records file byte counts, speed, the source site’s segmented responses and a bitmap of finished chunks. This is used only to display progress and resume tasks on your computer, and is not uploaded. Opening the details view only reads a local snapshot of the task.

What data the extension handles

This data is used only for the media detection, downloading, preview, conversion and related settings you can use in the interface. The extension does not use it for advertising, profiling or credit decisions, and does not sell it.

Where the data is stored

The media catalog and the general media cache are kept mainly in Chrome’s extension session storage; download tasks and some tool options are kept in the extension’s local storage on your computer. The platform media catalog keeps at most the 20 most recent entries, and the download-task index keeps roughly the 50 most recent tasks; these index limits are not a promise that all history is deleted automatically. Temporary storage is cleared when the browser session ends, and local records are removed when you clear them or uninstall, subject also to how Chrome behaves.

Some settings use chrome.storage.sync. If you have Chrome sync turned on, these settings can be synced between your browsers through Google’s sync service. Advanced service configuration uses the same storage, so do not keep credentials in it that you would not want synced. The extension has no sync server of its own.

Finished downloads are saved to your Chrome download folder, usually in a “Ruxget” subfolder. Uninstalling the extension does not delete media files you have already saved.

When it connects to outside services

Douyin, Kuaishou and Xiaohongshu

When you click the Ruxget icon on a single video page of these sites, the extension uses Chrome’s temporary “current tab” permission (activeTab) to read the video information the page’s player has already loaded: title, author, duration, cover image and the video address for each quality. It does not read or send cookies or account information, and it does not run unless you click.

The video servers of Douyin and Xiaohongshu require requests to carry the referring page. The first time you click download on either site, the extension asks for optional permission to reach their video servers (*.douyinvod.com, *.douyincdn.com or *.xhscdn.com). It is used only to add the current video page’s address as the referrer on those video requests and to download the video you chose. You can revoke it at any time on Chrome’s extension details page. Kuaishou does not need this permission.

The optional Ruxget desktop app

When you click “Desktop download” on a media item, or choose “Download with desktop app” as the download method and then click download, the extension asks for the optional nativeMessaging permission and, through Chrome’s and Edge’s native messaging, hands the Ruxget desktop app on this computer the selected download URL, title, quality, audio and video track addresses, and the Referer and User-Agent it needs. This video route does not forward cookies, Authorization or passwords, and it does not start the desktop app automatically when you install the extension or just read a page. General media that carries such authentication headers should use the original download route. The cloud-drive route described below does pass on the selected drive’s download cookies when needed.

The desktop app connects to the source sites and media CDNs you choose. By default it follows the Windows system proxy setting (it connects directly if no system proxy is on), and you can switch to no proxy or a custom address in the desktop settings. The download engine’s local RPC binds only to the loopback address and uses a random token. The browser does not connect to this RPC directly, and nothing connects to a developer server.

BT and magnet links (desktop app 0.4.61 and later): ordinary (non-private) torrents additionally use a public tracker list to find sources; magnet links that carry only a hash use it too while fetching the torrent’s metadata, and magnet links that bring their own trackers get the list added only once the torrent is confirmed not to be private. As with any BT download, trackers and DHT nodes can see your IP address and the hash of the resource. At most once a day the desktop app downloads this list from a public tracker-list host (cdn.jsdelivr.net, with raw.githubusercontent.com as a fallback); the request contains none of your information or downloads, and the list is stored on your computer. Private torrents never use public trackers. If your network (for example the “fake IP” mode of a router or proxy software) resolves tracker domains to fake addresses such as 198.18.x.x, then from 0.4.62 the desktop app looks up those trackers’ real addresses over HTTPS DNS (Cloudflare 1.1.1.1, with Google 8.8.8.8 as a fallback); the query contains only the tracker’s domain name, nothing about you or your downloads. Ordinary networks do not send these queries.

When you run the installer or re-pair on purpose, the program registers, for the current Windows user, the allowed extension ID, the local program path, the connection protocol and the uninstall entry; it does not scan your browser configuration. A connection check sends only a local ping and creates no media task. Provided local communication is already allowed, opening the download-method menu or the desktop setup and connection page makes the extension send the same ping to the desktop app to show the connection status; the ping contains no page, video or account information. A pairing request from an unknown extension ID is put to you on the desktop. The first video task downloads pinned-version components from the official Node.js and yt-dlp releases and the Gyan FFmpeg release; those download services can see ordinary network request information, and the download addresses do not contain your video links. Each component is checked against a fixed size and SHA-256 before it is installed and run, and no remote code list is accepted. The component cache is stored on your computer in data/components and can be paused and resumed. Uninstalling the desktop app keeps your tasks, unfinished data and finished files. Desktop tasks, source links, temporary signed addresses, settings and unfinished data are stored in the data folder next to the desktop program; finished files go to the folder you chose. You can close the program first and then inspect and delete this local data yourself. Uninstalling the extension does not delete the desktop app’s records or downloaded files.

Built-in downloads, turned on per cloud drive

Only after you open Ruxget on a specific cloud-drive page and click “Turn on Ruxget downloads” does the extension ask for the optional permissions for that drive’s pages, API domains, login cookies and local communication. New-style 123 shares include https://*.share.123pan.cn/*; requests are limited to the quota check and download-information paths on the current share domain, and no arbitrary API is opened up. Once authorized, a separate button appears next to the original download button; it does not replace the original button, does not take over Chrome downloads automatically, does not need Tampermonkey, and loads no remote scripts. Opening a cloud drive or the extension panel does not start the desktop app.

When you click “Ruxget download”, the extension reads the files you ticked and the login state it needs from the cloud-drive page, and asks the drive’s own API for download addresses. The request headers it needs are used only for allow-listed APIs, and the temporary network rules are cleaned up when the request ends. The file name, size, identifier, download address and the User-Agent and Referer it needs are handed to the desktop app through local communication. Quark and UC downloads also carry the authorized source site’s cookies, including HttpOnly cookies. Nothing is passed to the developer or to any third-party parsing service. The drive’s API and download service receive the access requests and the authentication information they need, as they normally would.

The desktop app adds the task to the save folder it remembers, without bringing up its main window while it receives the task in the background. The cookies and full signed addresses in cloud-drive tasks are encrypted with the current Windows user’s DPAPI and stored on your computer; the source address in the details view hides its query parameters so signatures and tokens are not exposed on screen. Cookies are cleared when the task finishes; the full address is kept with the task so it can be resumed, and Ruxget no longer uses it once the task is deleted. The local rotating backup may keep the previous encrypted record, and later saves update the backup. During a download, cookies are sent only to the HTTPS domains the drive allows, and a cross-domain redirect does not carry the original domain’s cookies. So the download engine can keep using segments and resume, the desktop app sets up, when needed, a temporary relay that binds only to 127.0.0.1 on a random address and accepts only file GET/Range requests; it offers no browser RPC and does not listen on the internet. When a signed address expires, go back to the drive to get a new one.

The cloud-drive route uses only the login state the web page already has (a cookie, or a login token the page keeps), needs no separate authorization of any third-party app, and does not send your login information anywhere except to Ruxget on your own computer. When you choose a save folder or view tasks, the relevant page or window opens as you act.

You can turn the route off in the extension panel for each drive, or revoke the optional site, cookie and local-communication permissions, to stop any further reading and handoff. Hiding the button only hides the entry point; downloads already handed to the desktop app must be paused or deleted in the desktop task list.

Original browser features

  1. Source sites and media servers: reading and downloading need to connect to the video sites and media CDNs you choose, for example Bilibili and its media servers. Requests carry the information the service needs, such as the video ID and referrer header, and the browser may attach login cookies under the original site’s rules. The target service also learns your IP address and request details through the normal network connection, and how it handles them is governed by its own policy.
  2. Chrome sync: if you turn sync on, Google processes the extension settings synced through Chrome.
  3. Advanced external features you use yourself: the settings and advanced tools keep optional Aria2, user-configured HTTP endpoints, MQTT, external players and online conversion / streaming-save pages. The default Bilibili MP4/MP3 flow does not depend on any of these services. Once you enable, configure or actively use one, the related media URLs, titles, request information, credentials or media content may be sent to the service you selected. The preset online pages include ffmpeg.bmmmd.com, ffmpeg.94cat.com and stream.bmmmd.com; the pre-filled MQTT address is test.mosquitto.org, and it is off by default. Send data only to services you trust, and avoid sending private URLs or credentials to a public message server.
  4. Links you click: after you open the Ruxget website or any other external link, that site handles the visit under its own policy.

The extension does not sell user data and does not use it for any purpose unrelated to the features above. User data is not used to determine creditworthiness or loan eligibility. Use of user data complies with the Chrome Web Store User Data Policy, including its Limited Use requirements; describing an optional external feature does not mean the service behind it has passed an independent privacy review.

How to stop, clear or delete

Under “Manage site permissions” at the bottom of the main help page you can revoke optional authorizations site by site; after that, further capture stops and the matching media cache is cleared. Old scripts already injected into a page exit when you refresh or close it. If an external service needs access, add its specific address on that page. Turning off media detection stops new general media capture; downloads already started must be cancelled on the task page. You can empty the media list and use “Clear captured data” on the settings page to clear the local cache; “Reset all settings” also clears synced settings. To remove all extension data, uninstall Ruxget and check your own Chrome sync settings. Files you have already downloaded must be deleted from your download folder yourself.

When you use advanced features, switching off the relevant toggle stops further sending; for data already sent to a third-party service, use that service’s own deletion process.

Contact and updates

For privacy questions and deletion requests, email the maintainer at support@hidaydream.com. Please do not send cookies, tokens, private URLs or other sensitive information unless they are needed to resolve your request.

Notes on individual features

Taking over Chrome downloads (0.4.51, off by default): only after you turn on “Take over Chrome downloads with Ruxget” on the “Desktop app setup and connection” page and allow local communication does the extension handle newly started ordinary downloads. It first pauses Chrome’s copy of the download and hands the link, the referring page address, the browser identifier and the file name and size Chrome reports to the desktop app on this computer; only once the desktop app confirms it can download directly is Chrome’s copy cancelled, otherwise Chrome resumes the download itself. Cookies, Authorization and form content are not passed on. During the handoff the extension uses the downloads.ui permission to hide Chrome’s own download bubble briefly and restores it right after the handoff (0.4.65); files left with Chrome show up as usual. Once takeover is on, ordinary files saved with “Download” in the extension panel are handled the same way (0.4.64); and when you click “Desktop download” in the panel while Chrome is already downloading the same link, that download is handed to the desktop app and Chrome’s copy is stopped, so nothing is downloaded twice. Downloads started by other extensions, downloads in incognito windows, temporary blob/data files, files smaller than the size you set, and sites you list are not taken over. Settings are stored on your computer; the 8 most recent takeover results (file name, link, whether it went to the desktop app or stayed with Chrome, the reason, and how long after the download started the extension stepped in) are stored only in your local browser to explain the reasons on the settings page, are not uploaded, and are removed when you uninstall the extension. You can turn it off at any time on the same page.

Save-location sync (0.4.52): the extension learns Chrome’s download folder from the file paths it has saved, keeps it on your computer, and sends it along when handing a download to the desktop app so the app saves to the same place (once you pick a folder in the desktop app, it stops following). This path is passed only between the extension and the desktop app on your computer and is not uploaded.

Desktop progress (0.4.54): once local communication is allowed, the extension asks the desktop app on your computer for the title, status, progress and speed of the tasks handed to it, to show them on the extension icon and in “Downloaded”; when the desktop app is not running, it only reads the task list the app keeps on your computer and does not start the app. A delete request is sent only after you click to confirm. The task IDs, file names and times of tasks handed to the desktop app (up to 20) are kept in the extension on your computer and are not uploaded.

Desktop app suggestion (from 0.4.56, replacing the takeover prompt of 0.4.54; 0.4.57 and 0.4.58 adjusted the pacing): when takeover is off, the extension reads the file name, size, progress and address of ordinary Chrome downloads, only to tell whether one is a large file of 50 MB or more that has not finished yet. The first time it meets such a download, the extension opens the Ruxget panel once under the toolbar icon and shows a card; it opens only after you have chosen a save location and the file has started downloading. Every successful automatic opening counts once (whether or not you answered); after that it waits 7 days, then 30 days, and requires 3 more large-file downloads in between, for at most 3 times in total. After you click “Don’t remind me” or turn takeover off yourself, it no longer opens automatically. The extension keeps on your computer only the download ID, file name, size and address behind this card (which expire after 24 hours), plus “how many times it has opened, when it last opened, how many large files came after, and whether you asked not to be reminded”; none of this is uploaded. Once local communication is allowed, opening the panel sends one ping to the desktop app on your computer to check whether it is installed; only after you click “Switch to the desktop app” are that download’s address, referring page, browser identifier and file name handed to the desktop app.

Main-video detection (0.4.54): when you open the Ruxget panel, the extension reads, in the current page (in frames it already has access to), the player’s size, whether it is muted, whether it loops, its playback progress and its media address, to put the main video first and fold away ads and previews. This is used only on your computer, is not stored and is not uploaded.

Opening downloaded files (0.4.68): when you double-click a finished download in the extension or click “Open”, the extension asks the first time for the optional downloads.open permission and, once allowed, uses Chrome to open the saved file (for example to play a video or run an installer). It opens a file only when you click, never opens any file automatically, does not read file contents, and does not collect or send any data. Windows still applies its security check based on the “from the internet” mark Chrome puts on downloaded files. For files finished through the desktop app (0.4.69), double-clicking in “Downloaded” has the Ruxget program installed on your computer open the file from the desktop app’s own task list: no desktop window pops up, nothing goes over the network, and again it happens only when you double-click.

The “Downloaded” tab (0.4.69): the “Downloaded” tab of the extension panel lists recent downloads, including files in Chrome’s download list (also downloads not started through Ruxget; downloads in incognito windows are not listed). Through the downloads permission it already has, the extension reads each download’s file name, download address (used to show a name when there is no file name), size, start time, progress and whether the file is still there. This information is read and shown only on your computer and only while the panel is open; no data is collected or sent. Clicking “Remove record only” on a Chrome download makes the extension note that download’s ID on your computer and stop showing it in Ruxget, while Chrome’s own download list stays unchanged; clicking “Delete file” or “Stop and delete” deletes the file (or stops the download) and removes it from Chrome’s download list. “Remove record only” on a Ruxget website-video row also removes it from Chrome’s download list, and the file is kept.

Settings page and handoff to the desktop app (0.4.71): the gear at the top right of the panel opens Ruxget’s settings page. Your “Hand over to desktop app” choice (never / large files and long videos / all), the save location, the floating button, the detection switch and the language are stored only in your local browser storage. Once you choose to hand over to the desktop app, videos or files start downloading directly in the desktop app on your computer, and what is handed over is the same as described below under “Optional desktop entry for ordinary files”; to avoid handing over the same video twice, the extension remembers the video addresses it has handed over for the current browser session (for at most 6 hours, cleared when you close the browser). The “Download Ruxget desktop app” links on the settings page, the desktop-app info page and the help page open the download section of the Ruxget website with ?from=ext in the address, used only so the website can count visits that came from the extension.

Clipboard (0.4.71): the extension writes the current page’s address to the clipboard only when you click “Copy this page’s link”, and never reads the clipboard. The desktop app reads the clipboard once, only at the moment you click “New download”: if it happens to hold a URL or a magnet link, it is filled into the new-download box for you to confirm; anything else is not read, stored or sent.

If the way data is handled changes, this notice will be updated, and the changes will be explained as the Chrome Web Store requires.

Optional desktop entry for ordinary files: only after you click “Download this link with Ruxget desktop app” in a link’s right-click menu, or confirm sending manually, is that HTTP/HTTPS link, the referring page address and the browser identifier handed to the companion program on your computer; cookies, Authorization and form content are not passed on. No ordinary download history is kept, and the clipboard is not read automatically.

When the desktop app is not connected, the connection guide appears only after you choose desktop download. Choosing to download and install opens the desktop-app page on the Ruxget website (hidaydream.cloud) (from 0.4.74), and the website records one visit “from the extension”, with no extension data included; repairing the connection opens the connection help inside the extension; choosing to carry on with Chrome, or cancelling, does not download, install or start the desktop app. No install pop-up is injected into the current web page.

Downloaded and Bilibili quality (0.4.72): “Downloaded” shows only downloads since this Chrome launch, in line with Chrome’s own download bubble; the extension notes the launch time only in browser session storage, cleared when Chrome closes, and earlier downloads are in Chrome’s own downloads page. When reading a Bilibili video, the extension asks Bilibili, from within the Bilibili page, whether you are currently logged in; when reading the list of parts (0.4.73), the extension’s background does the same query to Bilibili. The result is used only to explain that “Bilibili offers only lower qualities when you are not logged in”; it is not stored and not sent anywhere else.

Desktop auto-update (0.4.72): when “Download settings → Auto-update” is on (it is on by default), the desktop app reads a public version list from hidaydream.cloud at most once a day; when a new version exists it downloads the installer from the same website, checks its size, SHA-256 and digital signature, and installs it silently when you quit Ruxget or at its next start, keeping your tasks, settings and shortcuts. The request carries only the program name and version (Ruxget/version), with no download history, file names, account or device information; the website server records the request’s IP address and time, as it would for any page visit. When the installer is downloaded, the website also records a download entry without an IP address (time, file, whether it was downloaded completely, marked as an auto-update), used only to count auto-updates (0.4.73). With the switch off, these requests are no longer sent.

Video covers in the task list (0.4.73): the desktop app downloads the video’s cover image from the image server of the site the video is on (for Bilibili, Douyin, Kuaishou and Xiaohongshu the extension hands over the cover address together with the task; for videos the desktop app resolves itself, it fetches the site’s public thumbnail by video ID), shrinks it and shows it in the task list. Covers are stored only on your computer and deleted along with the task; the request carries no cookies or account information.

Buy me a coffee

Thank you. Support is entirely optional and changes nothing in the apps.

In WeChat, press and hold the code to scan it. In other browsers, save the image and scan it with WeChat.