You found an episode you want to reference later. Maybe it's an interview you want to quote, a long conversation you don't have time to re-listen to, or something you need to find a specific moment in. Getting the actual text out of a Spotify episode is not as obvious as it should be.
Here's how to do it.
Why Spotify doesn't give you the text
Spotify added transcripts to its app a couple of years ago, but there's a catch: you can't copy or export them. You can scroll through the text while listening, but the moment you close the app, it's gone. No download button, no "copy all." For anything beyond casual in-app reading, Spotify's native transcript is basically useless.
The workaround is transcribing the audio yourself using an AI tool — which takes about as long as making a coffee.
The fastest way: paste the URL into Podtyper
You don't need to download anything. You just need the episode URL.
On desktop: right-click the episode → Share → Copy Episode Link
On mobile: tap the ⋯ menu → Share → Copy Link
The link looks like this:
https://open.spotify.com/episode/4rOoJ6Egrf8K2IrywzwOMk
Go to podtyper.com, paste the URL, hit Transcribe. Podtyper grabs the audio directly — no file downloads, no extra steps.
Processing takes a few minutes regardless of episode length. A 3-hour interview and a 20-minute news update take roughly the same time because the transcription runs in parallel across the audio.
When it's done you get the full transcript with each speaker labeled and color-coded, a summary of the whole episode, key takeaways, notable quotes, and export options for TXT, SRT, or VTT.
Free tier: 30 minutes per month, no credit card. Start here →
Which Spotify episodes work
Almost all of them.
Any publicly accessible podcast episode works — full episodes, bonus episodes, trailers, episodes in any language. What doesn't work: Spotify-exclusive content that requires a premium subscription to access at all (rare), private podcast feeds, and Spotify audiobooks.
If you can listen to it for free on Spotify without logging in, Podtyper can transcribe it.
What people actually use Spotify transcripts for
Research and note-taking. You heard something important and want to find it again, or you want to quote it accurately. With a transcript you Ctrl+F the whole episode in seconds instead of scrubbing through audio.
Show notes. If you produce a podcast, transcription turns a 45-minute episode into show notes in about 10 minutes. Use the summary as the base, pull timestamps from the transcript.
Content repurposing. A single episode transcript is raw material for a LinkedIn post, a Twitter/X thread, a newsletter, and a blog post — all at once. This is how small teams produce more content than they should be able to.
Quoting accurately. Paraphrasing from memory is risky. A transcript gives you the exact words, timestamped.
Accessibility. If someone on your team is hard of hearing, or you need to consume content somewhere you can't play audio, having the text version is just more practical.
SEO for podcasters. Google can't index audio but it can index every word of a transcript. Podcasters who publish transcripts consistently see meaningful increases in organic search traffic within a few months.
How speaker labels work
Podtyper uses speaker diarization — the model listens to changes in voice, tone, and pacing to separate who's talking and when. Each section of the transcript gets attributed to Speaker 01, Speaker 02, and so on.
You can rename these directly in the transcript editor before exporting. So "Speaker 01" becomes the host's actual name, and that carries through to any export format.
Works well on standard interview-format podcasts, panel discussions with 3-4 speakers, and solo hosts with pre-recorded guest audio. Gets harder when two speakers have very similar voices, when there's a lot of crosstalk, or with 5+ speaker panels.
Accuracy
Podtyper uses Deepgram Nova-3, which handles conversational speech and accents well. On typical podcast audio — a host and one or two guests, recorded decently — accuracy is around 99%.
Where errors cluster: proper nouns (brand names, person names), technical vocabulary in niche fields, and moments where two people talk at the same time. A quick scan for these after you get the transcript is good practice. For most use cases the raw output is already clean enough to use.
Alternatives
Spotify's built-in transcript — works for in-app reading on shows that have it. Not exportable.
Otter.ai — you'd need to play the audio from your phone into Otter's microphone, which is clunky and loses quality. It doesn't accept Spotify URLs.
Whisper (OpenAI) — highly accurate and free, but you need to download the audio file, run Python locally, and process the output. Good if you're comfortable with a terminal.
Rev — human transcription, around $1.50/minute. Excellent accuracy but a 90-minute episode costs over $100 and takes 24-48 hours.
Podtyper is the only tool that takes a Spotify URL directly without any file downloads or setup.
Frequently asked questions
Do I need a Spotify account?
No. Podtyper accesses public podcast audio independently. You need a Podtyper account (free) but not a Spotify account.
Can I transcribe a full podcast series at once?
Each episode is transcribed individually. Batch transcription isn't supported yet.
What file formats can I export?
TXT, SRT, and VTT. All three download instantly from the transcript page.
Is the transcript stored privately?
Transcripts are stored encrypted and tied to your account. The shareable link feature lets you share specific transcripts publicly if you choose — by default they're private.
How does the summary work?
After transcription, Podtyper runs the text through an AI to produce a summary, key takeaways, and notable quotes. This happens automatically — nothing extra to do.
Getting a Spotify episode transcript takes about 5 minutes total: copy the URL, paste it into Podtyper, click Transcribe, come back when it's done.
Spotify's built-in transcript is fine for following along while you listen. For anything that requires actually using the text, you need a proper export.