Podcast Scraper: RSS Episodes, Transcripts & New-Episode Alerts

Podcast scraper for any public podcast RSS feed, by feed URL or by show name. For each show you get one clean row with the show's details, how often it publishes, whether it is still active, and its latest episodes with MP3 / audio links, transcript links and chapter links. Turn on Only new episodes and schedule it to get new-episode alerts.

No login, no API key. The Actor reads each publisher's own public RSS feed. It does not scrape Apple Podcasts or Spotify, does not download audio, and does not transcribe anything: transcript links appear only when the publisher includes them in the feed.

What can you use this podcast scraper for?

• Monitor podcasts for new episodes: schedule it daily and send only new episodes to Slack, email, Google Sheets or a webhook.
• Research shows: publishing cadence, days since the last episode and an active flag for a list of podcasts, e.g. for sponsorship or guest-booking research.
• Build an episode database or feed an AI pipeline: titles, dates, durations, descriptions, audio URLs and (where published) transcript links as flat JSON.

What you get

Per show

• title, author, website, language, categories, explicit flag, cover image URL, short description
• feed URL (after redirects), newFeedUrl if the feed announces a move
• episodes in the feed, latest and oldest episode date, days since the last episode
• publishing cadence: median days between the newest 21 episodes
• active flag: a new episode within the last 90 days (you can change this) and the feed is not marked complete
• podcastLocked (podcast:locked), podcast GUID, funding links (podcast:funding)

Per episode (newest first, nested in the show row)

• title, publish date, duration in seconds, audio URL, type and size, episode page link
• season and episode numbers, episode type (full / trailer / bonus)
• transcript links (podcast:transcript, all formats the feed offers) and chapters URL (podcast:chapters)
• GUID, plain-text description (HTML removed, length you choose)

Transcripts and chapters appear only if the publisher puts them in the feed. Many big shows don't.

Want one row per episode instead? Open the Episodes view in the Output tab, or add ?view=episodes&skipEmpty=true to the dataset items API URL. Each episode row also carries the show title and feed URL.

How to use it

1. Paste feed URLs into Podcast RSS feed URLs, and/or show titles into Podcast names.
2. Set Max episodes per show (default 10; 0 = show details only).
3. Optional filters: only episodes since a date or from the last N days, keywords an episode must or must not mention, skip trailers and bonus episodes.
4. Run. Export as JSON, CSV, Excel, or use the API.

Finding a show by name

Names are matched against a built-in index of about 260,000 popular podcasts, built from the Podcast Index public data dump (snapshot date in the run summary). The best match is used; each row shows nameMatch.matchedTitle, whether it was an exact title match, and up to 3 alternatives with their feed URLs. Common titles can match the wrong show, so for important shows, use the feed URL. Brand-new or very small podcasts may not be in the index yet.

New-episode alerts (monitor mode)

Tick Only new episodes since last run and schedule the Actor (for example daily).

• The first time a show is seen, its current episodes are remembered and nothing is returned for it (untick First run only saves a baseline to also get the newest episodes on that first run; the rest of the back catalogue is still remembered, so later runs return only episodes published after it).
• After that, only episodes not returned before come back. Shows with nothing new produce no row and cost nothing.
• If more new episodes appear than Max episodes per show, the rest come on the next run.
• Memory lives in a named key-value store in your account (podcast-tracker-state). Use State key to run several separate monitors.

Example input

{
  "feedUrls": ["https://feeds.simplecast.com/54nAGcIl", "https://lexfridman.com/feed/podcast/"],
  "podcastNames": ["Hard Fork"],
  "maxEpisodesPerShow": 5,
  "episodesWithinDays": 30
}

Example output (one show, shortened)

{
  "status": "ok",
  "input": "Hard Fork",
  "nameMatch": {"query": "Hard Fork", "matchedTitle": "Hard Fork", "exactTitleMatch": true, "alternatives": [{"title": "Hard Fork Café", "feedUrl": "https://anchor.fm/s/4a793c64/podcast/rss"}]},
  "title": "Hard Fork",
  "author": "The New York Times",
  "feedUrl": "https://feeds.simplecast.com/l2i9YnTd",
  "language": "en",
  "categories": ["Technology"],
  "episodeCountInFeed": 222,
  "latestEpisodeDate": "2026-10-30T10:00:00Z",
  "daysSinceLastEpisode": 0.2,
  "medianDaysBetweenEpisodes": 7.0,
  "active": true,
  "episodes": [
    {"episodeTitle": "Are We in an A.I. Bubble? We're About to Find Out.", "publishedAt": "2026-10-30T10:00:00Z", "durationSeconds": 3261, "audioUrl": "https://…/default.mp3", "episodeNumber": 219, "episodeType": "full", "transcripts": [], "chaptersUrl": null}
  ]
}

A run summary (shows, episodes, errors, monitor status) is saved as OUTPUT in the run's key-value store.

Broken or unreachable feeds

A feed that returns an error (404, 500, timeout, not a podcast feed, Atom instead of RSS, larger than 40 MB) gives a row with status: "error" and the reason. A name that can't be found gives status: "not_found". An input that leads to a feed already returned in the same run gives status: "duplicate". These rows are free, and the run carries on with the other shows. Slightly broken XML is read as far as possible and flagged in feedWarnings.

Pricing

Pay per event, no subscription. The start fee is only $0.00005 per run ($0.05 per 1,000 runs):

Event	Price
Actor start	$0.00005 per run
Show returned	$0.002 per show row
Episode returned	$0.0003 per episode ($0.30 per 1,000)

Error, not-found and duplicate rows are free, and in monitor mode shows with nothing new are free. Each show is charged once per run: names that differ only in case or spacing count once, and a name or URL that leads to a feed already in the run gets a free duplicate row. If you set a maximum charge per run, the Actor stops cleanly at that limit and says so in the status and in the row (episodesTruncatedByChargeLimit). Delivery is crash-safe: if the platform restarts or migrates a run, shows already delivered are not delivered or charged again (and error rows are not repeated).

Good to know

• Row size: a dataset item can be at most 9 MB. If a show's episodes (with long descriptions) would exceed about 8 MB, the row stops there, episodesTruncatedByRowSize is true and moreEpisodesAvailable is true. Only returned episodes are charged. Lower Description length to fit more.
• Big feeds are read one or two at a time to keep memory low; downloads still run in parallel.
• Data is only as good as each feed. Some publishers leave out durations, episode numbers or dates, and some feeds only list their newest episodes, so episodeCountInFeed can be lower than the show's real total.
• Times are UTC. "Days since last episode" is counted from the run time.
• Feeds are downloaded politely: at most 2 at once from the same host, with an honest User-Agent, timeouts and a short retry on server errors.
• No personal contact data. The Actor does not collect owner names or email addresses (itunes:owner, managingEditor, webMaster, the podcast:locked owner email) or podcast:person entries, and email addresses inside show notes are replaced with "[email removed]". author is the public show byline from the feed.
• Audio files are not downloaded; you get their links.

Source and credit

Show and episode data comes from each publisher's public podcast RSS feed (feedUrl in every row); content and rights stay with the publishers. Podcast-name lookup uses an index built from the Podcast Index public data dump (podcastindex.org), which Podcast Index allows to be downloaded, stored and redistributed; this Actor does not call the Podcast Index API. This Actor is not affiliated with or endorsed by Podcast Index, Apple Podcasts or any publisher.

Use it from code, Zapier, Make, n8n or an AI agent

• API: start runs and read results through the Apify API or the Python and JavaScript API clients. The API tab on this page has ready-made code.
• No-code integrations: connect the results to Zapier, Make, n8n, Google Sheets, Slack, email or a webhook with Apify's integrations.
• AI agents (MCP): add this Actor as a tool through the Apify MCP server, e.g. https://mcp.apify.com?tools=stego-bot/podcast-tracker. Input fields and output fields are fully described, so an agent can fill the input and read the results without guessing.
• Schedules: add an Apify schedule (e.g. daily) to turn a saved input into a recurring job or alert.

FAQ

Does it work with Apple Podcasts or Spotify links? It reads RSS feeds. Paste the show's RSS feed URL, or type the show name and let the built-in index find the feed.

Does it transcribe episodes? No. It returns transcript links only when the publisher publishes them in the feed (podcast:transcript).

Do I need an API key? No.

More Actors from stego-bot

• Bulk DNS Lookup & Domain Availability Checker
• EU Tenders Scraper (TED): Public Procurement Search
• SAM.gov Contract Opportunities API & Bid Alerts
• Greenhouse & Lever Jobs Scraper API + Job Alerts