SocialKit

SocialKit is a scraping API that turns YouTube, TikTok, Instagram, Facebook, X and LinkedIn URLs into clean JSON, including transcripts, AI summaries, key points, comments and engagement stats. It suits developers and research teams building social listening, influencer analysis or AI agent workflows, with MCP and no-code integrations.

SocialKit is a scraping and extraction API for social video and post data. It converts a public URL from YouTube, TikTok, Instagram, Facebook, X or LinkedIn into structured JSON, handling the proxy rotation, page parsing and platform quirks that would otherwise sit on your engineering backlog. For teams whose real need is video analysis, the interesting endpoints are the transcript and summarizer routes: a single request to the YouTube, TikTok, Instagram, Facebook or generic video-file API returns a written transcript plus a short summary, key points and detected topics. The product is aimed at developers and technical teams rather than casual end users. Anything that consumes social content at volume fits: a social listening dashboard, an influencer comparison tool, a research pipeline, or an AI agent that needs to know what a video actually said. Video files hosted outside the major platforms can be submitted through the dedicated transcript and summary endpoints, so internal call recordings and uploaded clips use the same schema as public posts. In practice, the workflow is deliberately thin. You pick an endpoint, paste the URL or send it programmatically with your access key, and receive a JSON payload with predictable field names. Responses are metered in credits, with most requests costing one credit and heavier queries (longer videos, more results) costing more. The homepage also documents a first-party Claude Code skill and a hosted MCP server, so an assistant such as Claude or Cursor can call the same endpoints as tools without custom glue code. Against alternatives, SocialKit competes less on raw scraping breadth and more on packaging: transcripts, AI summaries and engagement metrics arrive in one consistent response shape across six platforms, and no-code connectors remove the need for a bespoke scraper. The trade-off is that you are buying a data layer, not an interface: there is no visual analytics product here, and the value depends entirely on your own downstream tooling.