Use case
Podcast Transcription — turn episodes into show notes, blog posts and subtitles
One hour of podcast for €0.18, in 99+ languages, with automatic speaker diarization.
Podcasts are often a creator's most important content source, but audio alone doesn't index in Google. DeepScript transcribes episodes automatically with speaker diarization and word-level timestamps — so you can produce show notes, SEO blog posts and social snippets from one recording. A typical one-hour episode costs €0.18 and is ready in 2-5 minutes.
Recommended setup
- Model
- Premium (€0.27/hour)
- Language
- Set the episode language or use auto-detect. The Premium model is tuned for English and German conversation.
- Custom Vocabulary
- Add guest names, brand names, technical terms and product names to Custom Vocabulary — prevents "Stripe" being heard as "strike" or proper nouns getting mangled.
- Export
- TXT for show notes, SRT/VTT for YouTube/Spotify subtitles, JSON for your own pipeline.
How it works
- 1
Upload the episode
Drag MP3, WAV or M4A directly into the dashboard — full episodes over 1 GB are fine. The API works programmatically too.
- 2
Pick Premium + Vocabulary
Premium model for top accuracy. Add guest names and technical terms to vocabulary — saves a lot of cleanup later.
- 3
Review the transcript, pick an export
Quick scan in the editor and fix any typos. Then TXT for show notes, SRT/VTT for video subtitles, JSON for custom tooling.
Worked example
Example: 60-minute interview podcast with two speakers, Premium model, EN custom vocabulary with 12 terms. Cost: €0.27. Processing time: ~3 min. Output: full transcript with speaker labels + 90 word timestamps for SRT subtitles.
Frequently asked questions
How much does a typical podcast cost?
Standard model: €0.18/hour. Premium (better for conversation and guest names): €0.27/hour. For most podcasters the 9 cents extra for Premium pays for itself many times over in saved editing time.
Are speakers automatically separated?
Yes, in both tiers. If you know there are exactly 2 speakers ahead of time, you can set that as a hint at upload — improves separation on audio with background noise.
Can I create YouTube subtitles from this?
Yes. SRT and VTT export are included in every transcription — both formats are accepted by YouTube directly. Word timestamps make the timing accurate to 100 ms.
Does the audio stay with you or is it deleted?
Standard: automatically deleted after 30 days. With the Pro subscription (€22/month) audio + transcript stay permanently, accessible to AI agents via MCP.
Try it now?
Three transcriptions free, no credit card. Data stays in Germany. Three minutes from sign-up to finished transcript.