Add one MCP URL and Claude, ChatGPT or Cursor can read today's posts, ads, comments and transcripts from TikTok, Instagram, YouTube, Reddit and 12 more.
This project is scheduled for launch
Launch date: Sunday, August 1, 2027 at 08:00 AM UTC

LurkAPI is a social media data API. Send one GET request with query parameters and get clean JSON back from public pages on 16 platforms: TikTok, Instagram, YouTube, Reddit, Facebook, X (Twitter), LinkedIn, Threads, Bluesky, Telegram, Snapchat, Truth Social, Google, eBay, Linktree and any web page.
What you get: 88 endpoints for profiles, posts, comments, video transcripts, search, and the ads in Meta's, Google's, TikTok's and LinkedIn's ad libraries. One envelope everywhere: success, credits_charged, credits_remaining, then the data; errors carry a code and a link to the docs. Copy-paste code in curl, Node.js, Python, Go, PHP and Ruby on every endpoint page, with a real captured response and a live playground. An OpenAPI 3.1 spec, llms.txt, and every docs page as markdown.
For AI apps: the same endpoints are an MCP server. Add https://api.lurkapi.com/mcp to Claude, ChatGPT, Cursor or VS Code and sign in through the browser.
Speed and price: the API averages about a second. Most calls cost 1 credit, and packs run from $9 for 10,000 credits to $225 for 500,000. Credits never expire, and there is no subscription. Validation errors, rate-limited calls and 5xx errors are free.
Start free: sign in with Google or an email code and get 100 credits, no card. Once a day, a balance under 25 is topped back up to 25. LurkAPI returns public data only.
Comments will be available once the project is launched.