What your iPhone already does — and exactly where it stops
Since iOS 18 the phone transcribes on its own. Voice Memos shows a running transcript while you record, and Notes does the same for audio captured inside a note. It is good, it is free, and if that covers your case you should just use it.
Apple is precise about the boundaries, and they matter. Per Apple's own support documentation, audio transcription needs an iPhone 12 or later and covers ten language families — English, Spanish, Portuguese, Italian, French, German, Japanese, Korean, Simplified and Traditional Chinese — and it is "not available in all countries or regions."
The harder limit is the one Apple doesn't need to spell out: those features transcribe a recording you made on the device. A YouTube link, an episode someone else published, an mp3 on a server — none of those are recordings on your phone, and no amount of tapping in Voice Memos will reach them. That gap is the entire reason this page exists.
Three routes, honestly ranked
- Share sheet — fastest when you're already looking at the thing. You're in the YouTube app or a podcast player, you tap Share, you tap shrybe. No copying, no switching apps, no pasting. It costs about two minutes to set up once, because Apple only lets shortcuts be assembled on the device itself — the recipe is on our iOS Shortcut page.
- Safari — nothing to set up at all. Copy the link, open shrybe, paste, tap. Same engine, same transcript, zero installation. If you transcribe something once a month, stop here; the shortcut is not worth your two minutes.
- YouTube's own captions — free, but not a transcript. Many videos carry auto-generated captions you can read in the app. They are handy for following along and poor for anything you intend to quote, search, or paste elsewhere. Use them to skim; use a real transcript when the words have to be right.
What actually comes back
Separate paragraphs, one per segment the model detects, each stamped with its start time.
You can scrub to a moment, copy a quote with its timecode attached, or export the whole
thing. Plain .txt keeps the timings inline; .srt and
.vtt give you the same transcript as subtitle files — the phone handles all
three the same way a laptop does.
The engine is OpenAI Whisper, running the large-v3-turbo checkpoint on the free tier. Accuracy is not uniform across languages — it tracks how much of each language the model was trained on, which is a bigger spread than most people expect. We wrote up the per-language picture in how accurate Whisper actually is.
Where it won't help you
- Spotify. Streams are DRM-protected. No downloader reaches them, and a tool that claims otherwise is either wrong or doing something you don't want. Paste the episode from Apple Podcasts or the show's RSS feed instead.
- Private and members-only video. If a link needs a login to view, it needs a login to fetch. That is the platform's rule, not ours.
- Very long recordings. The free tier stops at 75 minutes per transcript. Most episodes fit; a three-hour stream does not.
- Heavy daily use. Fifteen transcripts per day, per address. Enough for real work, not enough to run a business on — the cap is what keeps the tier genuinely free.
Why the phone is the right place for this
Most transcription happens on a laptop because that is where the tools were built. But the moment you actually want a transcript is usually the moment you hear something worth keeping — mid-episode, on a walk, in a queue. That moment happens on the phone.
The reason it works well here is that your phone does almost none of the work. It sends a link; our backend fetches the audio from the source and runs it. A sixty-minute episode costs you a URL and a few kilobytes of text, not sixty minutes of cellular data or a drained battery. Lock the screen and the job carries on without you.
If you came for a specific source, those have their own pages: YouTube videos, Instagram reels, direct audio files, and subtitle files.
Frequently asked
Do I need to install an app?
No. shrybe runs in Safari — open the site, paste a link, read the transcript. The optional iOS Shortcut is not an app either: it is a four-action recipe you assemble once inside Apple’s own Shortcuts app, which then puts shrybe in the system share sheet.
My iPhone already transcribes voice memos. Why would I need this?
Apple’s built-in transcription works on audio recorded on the device — a voice memo you made, a recording inside Notes. It cannot reach a YouTube link, a podcast episode someone else published, or an mp3 sitting on a server. Apple also limits that feature to iPhone 12 and later and to ten language families. shrybe takes a URL and covers 100 languages.
How long can the audio be?
Up to 75 minutes per transcript on the free tier, which covers most full podcast episodes and conference talks. Longer than that, split the file or wait for the paid tier we are still designing.
Does it work on cellular, or should I be on Wi-Fi?
Either. The audio is fetched by our backend from the source URL, not from your phone, so your connection only carries the link and the finished text. That is also why a 60-minute episode does not eat 60 minutes of mobile data.
Can I keep using the phone while it transcribes?
Yes. Switch apps or lock the screen — the job keeps running on our side. Come back to the tab and the transcript is there; if the connection dropped in between, reopening the page recovers the finished result.
Can I get subtitles rather than plain text?
Yes. Every transcript exports as .srt and .vtt as well as plain .txt, with the timestamps intact. That works the same on the phone as on a desktop.
What about Spotify links?
Spotify streams are DRM-protected, so no tool can transcribe them directly and we will not pretend otherwise. Paste the same episode from Apple Podcasts, the show’s RSS feed, or its direct mp3 link instead — all three work.
The link is already in your clipboard.
Paste it above and read it in a minute. If you'd rather have shrybe in the share sheet first, the two-minute setup is here.