For podcasters

Your episode, transcribed in 60 seconds.

Paste an episode URL. Get a speaker-labelled transcript, an SRT ready for TikTok / Reels, and a DOCX show-notes block. First episode on us — no signup.

Direct mp3/m4a/wav URLs work best. Libsyn, Anchor, Transistor, Buzzsprout all supported.

Speaker labels built in ~60s for a 30-min episode Never trained on your audio 100+ languages

Built for the indie podcaster workflow

Three things most transcription tools get wrong for podcasts.

We fix them.

Speaker labels that actually work

Not "Speaker 1 / Speaker 2" guesswork. We diarize 2–8 speakers cleanly — guest interviews, co-host pods, panels. You fix the names once; the labels stay consistent through the episode.

SRT that's TikTok-ready

Caption cues at the right length — not 30-word walls of text. Drop the SRT straight into CapCut, Descript, Premiere, or Riverside. Word-level timestamps included if you need to re-cut.

Show-notes block, formatted

DOCX export includes a chapter-markers draft, key quotes pulled from the transcript, and timestamps linking back. Paste straight into Substack, Ghost, WordPress, Apple Podcasts.

Sample output

What you actually get back.

Speaker-labelled paragraphs, clickable timestamps, ready to paste into show notes.

episode-117.txt · transcript preview

HOST 00:00:12 Welcome back to the show. Today I'm joined by someone who's spent the last decade building podcasting tools from the inside out.

GUEST 00:00:22 Thanks for having me. I've been listening for a while so it's genuinely exciting to finally be on.

HOST 00:00:29 Let's start at the beginning. Back in 2016, you made what at the time was a pretty contrarian bet. Walk us through the thinking.

GUEST 00:00:41 The thing nobody talks about — most podcasters don't want a transcription tool. They want show notes that write themselves, clips that are ready for Reels, and a searchable archive of everything they've ever said. Transcription is just the input to that.

Every format you'd ask for

One transcript. Five clean exports.

.txt

Plain text

Clean, de-umm'd paragraphs. Drop into a blog post.

.srt

SRT captions

Social-clip-ready cues. Works with every editor.

.vtt

WebVTT

HTML5 player captions + YouTube uploads.

.docx

Show notes

Formatted with chapters + quote pulls for publishing.

Anything else

Questions podcasters actually ask.

How accurate is it on podcast audio?

We run Whisper large-v3 with word-level alignment. On clean 2-speaker podcast audio (Riverside / SquadCast / local recordings) we typically see under 5% WER. Heavy accents, music beds, and noisy field recordings hit harder — run a free trial to see for yourself.

How many speakers can you label?

Tested up to 8. Three-person panels and interview formats work cleanly. The first minute is usually enough to stabilise speaker identity across the episode.

What if my audio isn't online yet?

You can also upload directly at whipscribe.com — mp3, m4a, wav, flac, webm, mov, mp4 all work. Up to 2GB per file on the free tier.

Are my recordings used to train anything?

No. Your audio is transcribed and returned to you. We don't fine-tune on user content, don't ship it to third parties, and paid tiers get 365-day retention you can revoke at any time.

Do I need an account?

Not to try it. The first episode — up to 20 minutes of audio — is free to transcribe without signup. Sign in if you want retention, history, and API access.

I publish weekly — what does that cost?

A typical weekly podcast runs ~3-4 hours/month of audio. See the pricing page — Pro tier covers most indie creators; pay-as-you-go for one-off needs.