Why We Don't Ask You to Log In to Transcribe an Instagram Reel

Two logins people mix up

Most confusion around this question comes from two completely different logins being treated as one thing.

• Instagram login: gets you into Instagram. You need it to view a private Reel, save one, or use Instagram's own caption features. It never has to be handed to anybody else.
• Tool account: a sign-up on the transcription site itself. It does nothing for the transcript. Sites add it for history, quotas, billing, or syncing between devices.
• What transcription actually needs: the audio. One file on your device. Nothing about that file changes based on who you are.

What we checked on our own build today

This is not a policy statement. These are the measurements we took on our own endpoint on 2026-09-27.

• Two uploads of the same 61,200-byte MP3. One sent with no credentials at all, one sent with a fabricated Cookie: session_id=... and Authorization: Bearer ... header. Both returned HTTP 200 with the same transcript, beginning "Hello everyone! ...". The server has no credential to read, so a fake one changes nothing.
• Zero Set-Cookie headers on the home page, on an article page, and on both API responses. There is no session to establish.
• Zero uses of localStorage or sessionStorage in the home page HTML we pulled today (17,495 bytes). The only three occurrences of the word "login" on that page are our own copy saying there isn't one.
• One binding on the Worker: AI. No KV namespace, no D1 database, no R2 bucket. There is nowhere to write a user row even if we wanted one.

Where your file goes, and why an account would only add risk

An account would change that picture for the worse, not the better. Accounts need a credential store, sessions, and password reset flows — and once a site has an account for you, it has a natural place to keep a copy of your media. We have none of those pieces, so there is nothing to leak and nothing to subpoena.

No account is a design choice, not a missing feature

• The transcript is the deliverable. You copy it into your notes, your script, or your citation manager. Storing it for you would only create a second copy in a place you don't control.
• Your files already sync. They live on your phone and your laptop through whatever you already use. We would be re-solving a solved problem.
• No billing step between you and the text. No plan page, no trial timer, no card on file.
• No credential to leak. The strongest account security is not having accounts. It also means we cannot send you a breach notice, because there is nothing in the database.
• The same logic killed three other features. URL paste, because Instagram refuses server-side fetching on every method we tried. A browser extension, because it would sit between you and Instagram. An app, because the browser already picks a file fine.

What you need instead

• A file on your device. Our endpoint accepts MP3, WAV, M4A, AAC, OGG, FLAC, MP4, MOV and WEBM.
• Your own Reel: open it in Instagram and use the download or save option.
• Someone else's Reel: use Instagram's own save or share options where they exist, or screen-record it. The audio track is the part that matters for transcription.
• Long recordings: cut them into five-minute chunks first. In our own length tests, five-minute files succeeded three times out of three; fifteen- and thirty-minute files failed.

When an Instagram login genuinely is required

To see private or age-gated posts, to reach your own saved items, or to use Instagram's built-in caption tooling. Do those inside Instagram, where the login belongs. A transcription tool should never be the place where you type your Instagram password. If one asks, the reason is almost always that it wants to fetch media on your behalf — and that is the request Instagram refuses, with or without your password.