VidKeep

Guides

Saving Video as a Record

Portrait of Daniel Okafor, VidKeep engineer
Daniel OkaforVideo-tools engineer, VidKeep
· Published 2026-08-01 · Updated 2026-08-01 · 7 min read
Download result showing the source video and available qualities

The file is the easy part. What makes a saved video usable as a record is everything around it: where it came from, when you took it, and proof it has not changed since. A checksum, a timestamp and the original address take two minutes and are the difference between a record and a video file of unclear origin.

What to capture, beyond the video

WhatWhy it matters
The exact addressWhere it came from; the first thing anyone asks
Date and time you saved itEstablishes what existed when
Account name and display nameHandles change; both are worth recording
The caption or descriptionContext that is not in the file
A checksum of the fileProof it has not been altered since
A screenshot of the pageShows the video in situ, with counts and comments

Every row except the first is something that becomes impossible to reconstruct once the post is gone, and posts disappear constantly — deleted by the poster, removed by the platform, or taken with an account that closed.

The screenshot matters more than it looks. A video file shows the content; a screenshot shows that this particular content was published by this account on this page, which is a different claim entirely and frequently the one that matters.

Download result showing the source video title and available qualities
The download is one step; the context around it is what makes it a record.

Checksums, and what they actually prove

A checksum is a short fingerprint calculated from a file's contents. Change one byte anywhere in the file and the fingerprint changes completely and unpredictably.

On any Mac or Linux machine: sha256sum video.mp4. On Windows: certutil -hashfile video.mp4 SHA256. Both print a long string — record it alongside the file, with the date.

What that proves is narrow and useful: the file you have now is identical to the file you had when you recorded the fingerprint. It demonstrates the copy has not been edited since, which is the question people actually raise about a saved video.

What it does not prove is that the video is authentic, that the events in it happened, or that the platform served everyone the same file. A checksum is about integrity after download, not truth before it — and claiming more for it than that is the common mistake.

Worth being clear about, because it shapes how much weight a saved file can carry.

What you download is the platform's processed version, re-encoded on upload and compressed for delivery. It is not the original recording, and details that matter — a reflection, a background sign, a face at the edge of frame — may already have been lost to compression before you ever saw it — the general problem.

Take the highest quality available for anything you intend to keep as a record. The storage cost is trivial against the possibility of needing to look closely later, and unlike most downloads, here the extra detail is the point.

And keep in mind that platform video carries no reliable technical provenance. Camera metadata is stripped on upload; the file tells you almost nothing about where or when it was recorded, only about the encoding. Anyone hoping to establish origin from the file itself is looking in the wrong place — that information was removed before publication.

A workable routine

  1. Screenshot the page first, including the account, the caption and any visible counts. Before anything else, because this is what disappears fastest.
  2. Download the highest quality available.
  3. Verify the file is complete — skip to the end rather than watching the start — the check is here.
  4. Calculate the checksum and write it down with the date and time.
  5. Save the address, account and caption in a plain text file beside the video.
  6. Make a second copy elsewhere before you do anything else with it.

Six steps, about five minutes, and the order matters: the screenshot is first because a post can be deleted while you are downloading it.

If time is genuinely short — a post that looks likely to come down within minutes — the compressed version is screenshot, then download, and everything else afterwards. The two irreplaceable items are the picture of the page and the file itself; a checksum calculated an hour later still proves the file has not changed since then, which is most of its value.

Where this is genuinely appropriate

What connects these is that the material is public and the concern is that it will not stay available. None of them involves reaching content that was restricted, which remains impossible regardless of purpose — the honest answer on private content.

Storing a record so it survives

An archive kept badly is a record that cannot be produced at the moment it matters, which amounts to not having made it in the first place.

Keep the pieces together. The video, the text file, the screenshot and the checksum in one folder named for the subject and the date. Split across a downloads folder and a camera roll, they stop being a set within a week.

Two copies, in different places. A single drive is a copy waiting to fail — the longevity question in full.

Do not re-encode, ever. Converting the file to save space changes every byte and invalidates the checksum, which removes the one property that made it a record rather than a video.

Plain text for the notes. Not a document tied to software or a note locked inside an app that syncs somewhere. The point is that it opens in a decade without a subscription.

Limits worth being honest about

A saved video is weaker evidence than people assume, and understanding why prevents relying on it too heavily.

Anyone can produce a file. A downloaded video with a text file beside it is a claim, not a certification. What strengthens it is contemporaneity — a checksum and timestamp recorded at the time — and independence, meaning someone else captured it too.

Screenshots are trivially fabricated. They are still worth taking, because a screenshot plus a file plus a checksum plus a date is a coherent set that would take deliberate effort to fake, and no single element carries the weight alone.

The platform is the better source when it exists. Where a post is still live, a link to it beats any copy you hold. Your archive matters exactly when the original is gone, which is why it is worth making before that happens.

For anything with real legal stakes, this article is not the process you need. A lawyer or an archivist with a documented chain of custody is, and the difference between careful personal record-keeping and admissible evidence is substantial.

There is also a rights dimension that saving something for documentation does not dissolve. The video remains someone's work: keeping a copy to establish what was published is a defensible use, and republishing it is a separate decision with its own analysis — the four factors apply.

The practical distinction is between holding a record and distributing one. A file in your archive that supports an account of events is one thing; the same file posted publicly to make a point is publishing, whatever the motive, and it is judged as such.

What our download preserves, and what it cannot

Since this article is about provenance, it is worth being precise about what passes through our service unchanged and what is simply not available to preserve.

The video data is untouched. Streams are copied into a container rather than converted, so the picture and sound in your file are bit-for-bit what the platform served. Nothing is re-encoded, which matters here more than anywhere else on this blog: a re-encode would change every byte and make any comparison with another copy meaningless.

That has a useful consequence. Two people downloading the same video at the same quality should get files that match, because neither copy was altered in transit. Independent capture producing identical checksums is worth considerably more than one person's file, and it is only possible if the tools involved pass the data through rather than processing it.

What we cannot preserve is everything outside the video. We show the title and duration and discard the rest of what the platform tells us — account name, counts, identifiers — before it reaches any log. That is a privacy decision made for ordinary users and it means our service gives you nothing to build a record on. The caption, the screenshot and the address are yours to capture.

We also do not timestamp anything or hold a copy that could corroborate yours. Files are streamed through and never stored, and the event log keeps a hashed link with no title and no account — the full field list is published. If you needed us to confirm that you downloaded a particular video on a particular day, we could not, and that is the same design that means nobody else can compel it from us either.

Which is a trade worth naming rather than presenting as a pure benefit. A service that kept records could corroborate your account of what you saved and when — genuinely useful in the narrow case this article describes, and a database of who watched what in every other case. We chose the version that helps the majority and leaves the documentation work with the person doing it.

Frequently asked questions

What should I save besides the video?

The exact address, the date and time, the account name, the caption, a checksum of the file, and a screenshot of the page.

What does a checksum prove?

That the file has not changed since you recorded the fingerprint. It says nothing about whether the video is authentic or what it depicts.

How do I calculate one?

sha256sum on Mac or Linux, certutil -hashfile with SHA256 on Windows. Record the result with the date.

Which quality should I download for a record?

The highest available. Storage is cheap against the possibility of needing to look closely, and detail is the point here rather than convenience.

Is a downloaded video usable as evidence?

It is a claim rather than a certification. Contemporaneous checksums, timestamps and independent copies strengthen it; for real legal stakes you need a documented process.

Can you confirm that I downloaded something?

No. Files are never stored and the event log keeps only a hashed link with no title or account — which also means nobody can compel that information from us.

Try it yourself

VidKeep runs in your browser — paste a link, pick a quality, keep the file. No account, no app.

Open VidKeep

Last updated: 2026-08-01. We revise our guides as the platforms change.