Framer AI AgentsDesign and publish professional sites with AI
Promoted
PodKit
Podcast search, metadata & transcripts API — from $15/mo
9 followers
Podcast search, metadata & transcripts API — from $15/mo
9 followers
PodKit is a developer-friendly podcast data API: search, episode metadata, chapters, and transcripts as clean JSON — starting at $15/month, with caching allowed and no attribution logo required. Ships a first-party MCP server so AI agents (Claude, etc.) can use it directly as tools on the same key and quota.
Would love to see webhook support for new episodes on subscribed shows, since polling the API to check for updates feels wasteful. Even a simple push notification on new content would save a ton of quota for anyone building monitoring tools on top of this.
@budakcnuri97568 Great point — polling is wasteful for exactly that use case. Webhooks for
new episodes on subscribed shows are on my radar for the next tier. Would
love to hear more about what you're building if you want to share — happy
to keep you posted when it ships.
The fact that transcripts and chapters come back as clean JSON without any extra fluff is such a relief compared to wrestling with RSS feeds. Shipping an MCP server on the same key is a really thoughtful move for AI workflows.
@uur29784567603 Thanks! That was the whole point — wrap the RSS/chapters/transcript mess
so nobody else has to deal with it. Glad the MCP server made that easy for
agent workflows specifically.
would love a webhook option for new episode drops so my pipeline can react without polling the search endpoint every few minutes, would save a lot of requests on my quota
@cemgvezryjy yeah this keeps coming up — you're the second person to ask for exactly
this since launch. webhooks for new episodes instead of polling is now
top of my list for the next update. what's your polling interval right
now, out of curiosity — trying to get a sense of how much quota this
would actually save people
finally an API that doesn't make me jump through hoops just to get transcript data, and the MCP server actually worked on the first try with my Claude setup
@kazaz18975
That's great to hear — "worked on the first try" is exactly the bar I was
aiming for with the MCP setup. What are you building with it, if you don't
mind sharing?
finally an api that doesn't slap an attribution logo on my project, and getting transcripts back as json instead of scraping rss feeds saved me a weekend
@merve1533881 that's exactly the pain point I built this around — the RSS scraping
tax is real, especially once you factor in chapters/transcript format
differences across podcasts. glad it saved you the weekend