Transcribing Is Easy. Using the Transcript Is the Work.
You can transcribe a podcast episode in a couple of minutes now — getting text out of the audio is close to a solved problem. The reason most shows still don't publish transcripts or repurpose properly isn't transcription difficulty — it's that a 12,000-word wall of text isn't obviously useful, so it sits in a folder.
This guide covers both halves: getting the transcript, and turning it into something.
Getting the Transcript
If your episode is on YouTube — as most video podcasts now are — paste the URL into ReelWords. It looks for the existing caption track first, which is fast and free, and falls back to AI transcription if there isn't one.
If you have the file — an MP3, MP4, or MOV from your recording setup — upload it directly. Size limits apply by plan (25 MB free, 200 MB Pro), so long episodes may need the URL route or a compressed export.
If it's published on a social platform — a clip on TikTok, Instagram, or Facebook — public URLs from those platforms are supported too.
Export the format that matches what you're doing next:
| Format | For |
|---|---|
| Plain text | Writing show notes, articles, newsletters |
| Timestamped text | Finding clip-worthy moments by time |
| SRT / VTT | Subtitling video versions or clips |
For repurposing work, timestamped text is usually the right choice. You need to know *where* the good moment is, not just that it exists.
A Realistic Caveat on Multi-Speaker Audio
Most podcasts are conversations, and this is where transcription gets harder.
ReelWords does not do speaker diarisation — it won't label who said what. For a two-person interview, the transcript reads as continuous text, and you'll be inferring speakers from context.
That's fine for finding quotes and clips, where you're going back to the audio anyway. It's a real limitation if you want a publishable, speaker-attributed transcript. For that, a tool with speaker identification is a better fit, and it's worth knowing before you build a process around this.
Overlapping speech — people talking over each other, which happens constantly in good conversation — is also the hardest case for any transcription system. Expect to clean those passages up.
What to Do With the Transcript
This is where the value is.
1. Show notes. The transcript gives you the episode's actual structure. Pull the topic shifts, the key claims, and anything referenced, and you have your notes without re-listening.
2. Find your clips. Read the timestamped transcript and mark the strongest 30-60 second passages. This is dramatically faster than scrubbing audio, and you're reading for substance rather than trying to remember where something was. Take those timestamps back to your editor and cut.
3. Caption those clips properly. Podcast clips on social live or die on captions — most people watch on mute. Once you have the clip, caption it with word-level styling rather than flat subtitles.
4. Write the article. A good hour-long episode contains a solid 1,200-word article. Starting from the transcript means starting from your own actual arguments rather than a blank page.
5. Pull quotes for social. Direct quotes with attribution make strong standalone posts. The transcript gives you exact wording.
6. Make the archive searchable. Transcribe your back catalogue and suddenly "which episode did we talk about pricing on?" is answerable. For shows with 100+ episodes this compounds into something genuinely valuable.
7. SEO. Published transcripts are indexable text on a page where otherwise there's only an audio embed. Episode pages with transcripts have something for search engines to actually read.
Shortcut: Summarize Before You Transcribe
If your episode is on YouTube, there's a faster path for some of these jobs.
An AI summary returns key insights, actionable takeaways, memorable quotes, facts and examples cited, and the people and products mentioned — structured, in about a minute of reading.
That covers show notes and quote-pulling almost entirely, without reading 12,000 words. Pull the full transcript when you need precise wording or clip timings.
For a back catalogue, the pattern scales well: summarize everything, then transcribe selectively.
A Workflow for a Weekly Show
Something repeatable:
- Publish the episode to YouTube alongside your audio host.
- Summarize it — key points and quotes become the backbone of your show notes.
- Pull the timestamped transcript and mark three or four clip candidates.
- Cut those clips in your editor using the timestamps.
- Caption the clips with styled, word-level captions for social.
- Publish the transcript on the episode page for search and accessibility.
- Write one article from the strongest segment.
That's one recording session producing an episode, show notes, four social clips, a published transcript, and an article.
What It Costs
| Plan | Transcripts per month | AI summaries per month |
|---|---|---|
| Free | 10 | 3 |
| Pro | 100 | 50 |
A weekly show fits inside the free tier for transcripts. Heavy back-catalogue work or multi-show production wants Pro. See pricing.
FAQ
How do I transcribe a podcast episode for free?
If the episode is on YouTube, paste the URL into ReelWords — the free plan includes 10 transcripts per month. If a caption track already exists, retrieval is near-instant.
Can I transcribe an MP3 audio file?
Upload works for video and audio files within your plan's size limit. For long episodes, publishing to YouTube and using the URL avoids upload size constraints entirely.
Will the transcript label who is speaking?
No. Speaker identification isn't part of ReelWords. For publishable, speaker-attributed transcripts of interviews, use a tool with diarisation.
How accurate is podcast transcription?
Good on clean, single-speaker audio. It degrades on overlapping speech, remote guests with poor connections, and heavy background music. Review before publishing.
Should I publish my podcast transcript?
Generally yes. It's indexable content on an otherwise text-light page, and it makes the episode accessible to people who can't or won't listen.
What's the fastest way to find clips in an episode?
Read the timestamped transcript rather than scrubbing audio. Mark the passages that work standalone and take the timestamps to your editor.
Can I transcribe a whole back catalogue?
Yes, within your monthly limits. Many shows summarize the full catalogue first, then transcribe selectively based on which episodes matter.
Start With Your Last Episode
Pull the transcript, mark three clip candidates, and see how long it takes compared to scrubbing the audio.
That comparison usually settles whether this belongs in your weekly process.
Once the clips are cut, the caption styles handle the part that decides whether they get watched on mute. See pricing for monthly limits and the FAQ for workflow questions.