Privacy Policy

Reel is a self-hosted video downloader. This page describes, plainly, what data passes through the service while you use it and how long it sticks around — not a generic template, but a description of what the code actually does.

Reel doesn't have user accounts, sign-in, or a saved history of your downloads. Every visitor is anonymous. There is no page anywhere that lists videos you've previously downloaded.

On your first request, Reel sets an session_id cookie — a random identifier with no personal information in it, marked HttpOnly so page scripts can't read it. It exists for one reason: to apply per-visitor rate limits fairly, so one browser tab can't be mistaken for another. It isn't used for tracking across sites or for advertising.

When you start a download, Reel creates a temporary job record to track its progress. That record holds: the source URL you submitted, the format you chose, a status (queued/processing/completed/failed/cancelled), a progress percentage, the video's title and duration if known, the resulting file size once known, an error message if something went wrong, and timestamps for when the job was created, started, and completed. It also stores the requesting IP address and the anonymous session id, used together to enforce the concurrent-download and rate limits described below.

This record is metadata only — Reel never stores the video or audio bytes themselves in this table. The actual file lives briefly on temporary disk storage and is deleted once it's been delivered, once it expires, or once processing fails (see below).

The downloaded file sits in temporary storage only long enough for you to retrieve it. It's deleted immediately after your browser finishes downloading it. If you never retrieve it, an expiry sweep removes it (and the job record) automatically after a configured time-to-live window. Nothing is kept indefinitely by design.

Your IP address is used to apply rate limits (so the service stays usable for everyone) and a concurrent-download cap per IP (so one visitor can't queue an unlimited number of simultaneous jobs). It's associated with job records for this purpose and isn't shared or sold.

If the operator running this instance has configured it, unexpected server errors may be reported to Sentry for debugging. User-facing error messages shown to you are always generic and never include stack traces, file paths, or database details — see the service's exception handling, which is deliberately built to never leak internals in a response. Error tracking is optional and disabled by default.

Reel doesn't run advertising, analytics pixels, or third-party tracking scripts. The only outbound network activity from your browser is to this service's own API.

If how this instance handles data changes, this page will be updated and the effective date above will change accordingly.