Paste a Spotify or Apple Podcasts episode link and get the whole episode as timestamped text, with speaker labels if you want them. There is nothing to download: we find the episode in the show's public RSS feed and transcribe that audio. An hour-long episode takes about 1–3 minutes, and the first 30 minutes are free without an account.
Without signing in: the first 30 minutes. A free account: up to 72 minutes per recording, within your remaining credits. Longer recordings return a preview.
Compare plansmp3 · wav · m4a · flac · and more
Copy the share link in Spotify (open.spotify.com/episode/…) or Apple Podcasts. A direct link to an audio file or the show's RSS feed works too, and from a feed you can pick up to 10 episodes at once.
Tick Identify speakers for an interview or a panel, and each line is labelled Speaker 1, Speaker 2 and so on. Leave it off for a solo show.
Search the transcript, rename the speakers, summarize the episode, or download TXT, DOCX, SRT or VTT. PDF comes from the print dialog.
The one whose limits fit your episode. Five things are worth comparing: how much is free, the longest episode a plan takes, whether it labels speakers, which file formats you get, and the monthly price. Scribetape's answers are on this page, taken from the limits the product enforces, so you can set them against any other tool's pricing page. The quickest test is to paste your episode above: the first 30 minutes cost nothing.
Without an account you get the first 30 minutes of any episode, speaker labels included. A free account gets 5 hours a month, for episodes up to 72 minutes each.
Up to 3 hours on Essential, 6 hours on Growth and 24 hours on Pro, per episode. A free account takes 72 minutes. A longer episode isn't refused: you get the opening as a marked preview and can continue it on a plan that covers the rest.
One credit per started hour of audio, two with speaker labels. Essential is $6.99 a month for 100 hours, Growth $9.99 for 400 hours and Pro $19.99 for 1500 hours. A 2-hour episode uses 2 credits, or 4 with speaker labels.
Yes. Tick Identify speakers and each line starts with Speaker 1, Speaker 2 and so on. Rename them to real names on the transcript page and the names carry into every export. It counts double the hours and works on episodes up to 3 hours.
TXT, DOCX for Word or Google Docs, SRT and VTT subtitles, and PDF through the print dialog. Every format is on every plan, the free preview included, and the files carry no watermark. A preview that stops early says so in the file or in its name.
Whisper large-v3 Turbo, OpenAI's open speech model. Speaker labels are added by pyannote, which tells the voices apart. The episode is transcribed in the language spoken, detected automatically, and you can translate the transcript into 16 languages with the timestamps kept.
It does best on clear speech with one person talking at a time, and slips most on names, jargon and people talking over each other. Every line carries its timestamp, so you can check it against the audio. It is AI only: no person reads the transcript, so have it checked before it goes to court or into print.
About 1–3 minutes for an hour-long episode, because the audio comes as one file from the show's feed. You can close the tab; the link keeps working.
Any show with a public RSS feed, pasted as a Spotify, Apple Podcasts or RSS link, and any audio file you upload. Spotify-exclusive shows and subscriber-only episodes have no public audio to fetch. From a show's feed you can pick up to 10 episodes and transcribe them in one go.
Yes, once signed in: an overview, a timeline with timestamps and the key points, for 1 credit per hour of audio.