Privacy
What this site collects.
Short version: two analytics tools on this website, and an application that collects nothing unless you switch it on and cannot identify you if you do.
Last updated 31 August 2026
The website
Two analytics tools run on these pages, and they are the only things here that load from anywhere other than this domain.
Google Analytics 4
Property G-KNGWP21KK5. It sets cookies and sends page views, along with the usual coarse details — approximate location from your IP address, browser, device type, and the page you came from — to Google, where Google's own privacy policy applies.
Umami, self-hosted
Runs on a server we control. It is cookieless and records page views without building a profile that follows you between sites. Nothing it collects is shared with anyone.
Both are there on purpose rather than by accident. Google Analytics is blocked by a large share of the people who read a page like this, so running a self-hosted tool alongside it is the only way to know how much the first one is missing. The gap between the two numbers is the measurement.
There is no cookie banner. Google Analytics sets cookies and you are not asked first. That is worth stating rather than burying: if you would rather not be counted, any content blocker stops it, and Firefox and Safari block it by default.
What is not here
- No forms, accounts, or newsletter. There is nothing on this site you can submit, so there is nothing for us to store.
- No third-party fonts. The five font files are served from this domain, so no request goes to Google Fonts or any other host.
- No advertising, and nothing sold or shared.
The site is hosted on Sevalla and served through Cloudflare, both of which keep ordinary server logs of requests. Downloads are served by GitHub, so following a download link is a request to GitHub and their policy covers it.
The application
Flume collects nothing by default. It asks once, during first run, whether it may send anonymous usage counts, and sends nothing at all unless you say yes. Declining is permanent — it does not ask again.
Diagnostics report
Never sent anywhere. Builds a report you can paste into a bug report, shown on screen first so you copy it yourself. Flume has no way to transmit it.
Usage counts
Off unless you turn them on. Settings → Privacy → Send anonymous usage counts. Everything below describes what happens only after you do.
What usage counts contain
At most once an hour, and once when you quit, five fields go to a collector we run:
| installId | A random UUID made on your machine when you consented. Not derived from your hardware, network, or username — it has no relationship to you, so it cannot be linked back to you or matched against any other application. |
|---|---|
| appVersion | Which version of Flume, e.g. 1.0.0. |
| os | macOS, Windows, or Linux. |
| arch | x86_64 or aarch64. |
| events | Which things happened — launched, a torrent added, a torrent completed, a setting changed, an operation failed, and a handful more. |
Two things keep that from identifying anyone. Counts are sent as ranges rather than exact numbers, because a range is what anyone would graph and "the install with 1,483 torrents" is one identifiable person. And events are timed to the hour, not the second.
There is no free-text field anywhere in the format. Every value Flume can send comes from a fixed list, which is why which setting you changed is sent and what you changed it to is not.
What it never collects
- What you download. No torrent names, file names, info hashes, or magnet links.
- Who you download it from. No tracker addresses, peer addresses, or DHT nodes.
- Where you put it. No file paths, folder names, or drive names.
- Who you are. No account, email, name, machine identifier, hardware serial, or MAC address.
The collector is a Cloudflare Worker writing to a small database, and its source is in the repository — around 250 lines, short enough to read in full. It does not store your IP address or your user agent: the code never reads either, and the database has no column for them. Cloudflare sees your IP the way any web server does, but nothing we control writes it down.
Turning it off
Same switch, immediately, with no restart. The install ID and any queued events are deleted from your machine. What was already received cannot be tied back to you — the ID is gone from your side and was never linked to you in the first place — so there is nothing to request the deletion of.
One honest limitation, in the diagnostics report. It strips paths, addresses, URLs, and info hashes by pattern, but torrent names have no pattern — they are matched against your current library. A log line naming a torrent you already removed has nothing to match and can survive. That is exactly why the report is shown before it is copied: you are the only person who can recognise it. Read it before you paste it.
Everything Flume connects to
A torrent client cannot be silent on the network. This is the complete list: trackers listed in the torrents you add, peers (directly or through your SOCKS5 proxy), the DHT if you leave it enabled, your router if UPnP is on, and the usage collector only if you opted in.
There is no update checker, no crash reporter, no analytics SDK, and no font or asset fetching — the fonts are bundled, so the interface renders with no network at all. The full detail is on the app's privacy page, including every event and what it carries.
What BitTorrent exposes
This part is not about Flume, and no client can change it. It is worth saying anyway.
- Peers can see your IP address. That is how the protocol works — peers connect to each other directly. Anyone in a swarm can record who else is in it.
- DHT announces that you are in a swarm. You can turn it off in Settings; magnet links stop working when you do.
- Flume can route peer connections through a SOCKS5 proxy, which changes what peers see. It routes peer traffic and is not a general-purpose anonymity tool. If anonymity is your threat model, a torrent client is not the control you are looking for.
Questions
Email adamgreenwell@gmail.com, or open an issue on the repository for this site. If something on this page is wrong or out of date, that is a bug and it will be fixed.
This page describes what happens. It is not legal advice and it does not claim compliance with any particular regulation.