Audio, Video & Podcast Transcriber (Whisper Large v3)

Turn audio, video and podcast episodes into text and subtitles with Whisper large-v3 on fast GPUs. Paste file links or a podcast (Apple Podcasts link or RSS feed) and get the full transcript, timestamped segments and ready-to-use SRT and VTT subtitle files. 99 languages, auto-detected, plus translation to English. A 10-minute episode takes about 15 seconds.

What you can use it for

• Podcasters and creators: show notes, blog posts and searchable archives from every episode.
• Video editors: SRT/VTT subtitles for YouTube, Reels and courses, split into short two-line captions.
• Researchers and journalists: interviews, lectures, hearings and calls in any language, with timestamps.
• AI and automation: feed transcripts to ChatGPT, Claude or your RAG pipeline from Make, n8n, Zapier or the API.
• Media monitoring: schedule a daily run to transcribe the newest episodes of the shows you follow.

Input

Field	What it does
Audio or video file URLs	Direct public links: mp3, m4a, wav, ogg, mp4, webm.
Podcasts	Apple Podcasts show links or RSS feed URLs. The newest episodes are transcribed.
Episodes per podcast	How many of the newest episodes to transcribe per show (default 3).
Only episodes whose title contains	For example a guest name or a topic.
Spoken language	auto detects it, or set a code (en, hi, es, de, ar, ja...) for the best accuracy.
Task	Transcribe in the original language, or translate to English.
Subtitles	Creates SRT and VTT files for every transcript.
Max minutes per file	Longer files are skipped and not charged.

{
  "mediaUrls": ["https://your-site.com/interview.mp3"],
  "podcasts": ["https://podcasts.apple.com/us/podcast/the-indicator-from-planet-money/id1320118593"],
  "maxEpisodesPerPodcast": 1,
  "language": "auto",
  "subtitles": true
}

Output

One row per file or episode (real output, October 2026; text and segments shortened):

{
  "title": "Supercore inflation, Bumble valuation and Pooh's repatriation",
  "podcast": "The Indicator from Planet Money",
  "publishedAt": "2026-10-16T07:00:00.000Z",
  "episodeUrl": "https://www.npr.org/2026/10/16/nx-s1-6004278/supercore-inflation-bumble-valuation-and-poohs-repatriation",
  "durationSeconds": 644,
  "minutesCharged": 11,
  "task": "transcribe",
  "languages": ["en"],
  "text": "NPR. Waylon, is that an animatronic bird? Oh, why yes. This is Molly, my one-of-a-kind singing and dancing bird automaton. ...",
  "segments": [
    { "start": 6.6, "end": 28.4, "text": "Waylon, is that an animatronic bird? Oh, why yes. This is Molly, ..." }
  ],
  "srtUrl": "https://api.apify.com/v2/key-value-stores/xOxjtjjRfNiiK6cT4/records/transcript-0002.srt?signature=...",
  "vttUrl": "https://api.apify.com/v2/key-value-stores/xOxjtjjRfNiiK6cT4/records/transcript-0002.vtt?signature=...",
  "processingSeconds": 17
}

The subtitle files use short captions of up to two lines, ready to upload to YouTube or any video editor:

2
00:00:06,600 --> 00:00:10,942
Waylon, is that an animatronic
bird? Oh, why yes.

3
00:00:10,942 --> 00:00:16,880
This is Molly, my one-of-a-kind
singing and dancing bird automaton.

Pricing

Pay per audio minute: $5 per 1,000 minutes (that is $0.30 per hour of audio) on paid Apify plans, less on Scale and Business. Each file is rounded up to the next full minute. Files that fail or are skipped are never charged, and the run stops cleanly at your spending limit.

On the Apify free plan, each run transcribes up to 2 files of up to 20 minutes each. Any paid plan removes these limits.

Tips

• Set the spoken language when you know it. It is faster and more accurate than auto-detect, especially for short clips.
• Spotify shows have no public audio feed. Use the show's Apple Podcasts link or RSS feed instead.
• YouTube page links are not supported. Use a direct link to the audio or video file.
• Files over 4 hours are skipped by default. Raise Max minutes per file for very long recordings.