Social Fetch

適用於 TikTok、Instagram、YouTube、X 等平台的 Scraper API。

造訪 socialfetch.dev

文件

Documentation

Look up profiles, posts, and engagement metrics across every major social platform through one normalized REST API.

Use with AI

Social Fetch is a REST API for normalized social media data. Look up profiles, posts, and engagement metrics across Instagram, TikTok, X/Twitter, YouTube, and more — one JSON shape, one API key.

Start with your AI tool

Paste this prompt into Cursor, Claude Code, or any AI IDE. It will read the docs and integrate Social Fetch for you.

PreviewCopy prompt

The essentials

Base URL https://api.socialfetch.dev · authenticate with the x-api-key: sfk_... header · grab the spec from /openapi.json. New here? Start with the Quickstart.

Pick your path

Quickstart

Make your first authenticated request and learn the response envelope in a few minutes.

Recipes

Copyable workflows — brand monitoring, transcripts, ads, creator scoring, Reddit research.

TypeScript SDK

Typed client methods and Result-based error handling instead of raw HTTP.

API reference

Every operation, with parameters, examples, and response schemas.

Credits & billing

Which routes cost credits, how to read creditsCharged, and what 402 means.

Integrations

SDK, MCP, n8n, Apify, x402 — connect Social Fetch to your stack.

Not sure where to begin? Match your setup:

  • Calling the API directly — Quickstart → Errors → Credits → API reference.
  • Building in TypeScript — Quickstart → SDK guide.
  • Handing this to a coding agent — point it at /llms.txt, /agents.txt, /llms.json, and /openapi.json, or copy the prompt above.

Platform coverage

Every platform returns the same { data, meta } envelope.

  • Instagram — profiles, posts, reels, highlights, search
  • TikTok — profiles, videos, products, user search
  • X/Twitter — profiles, tweets, communities, search
  • YouTube — channels, videos, transcripts, playlists, search
  • Facebook, LinkedIn, Reddit, Threads, Telegram, Linktree, Hacker News, Truth Social, Spotify — profiles, posts, and platform-specific lookups
  • Web — fetch any page as Markdown or HTML, or ask a question about it
  • Utilities — whoami, balance, and health

See the API reference for every operation and its schema.

Docs for machines

Building with a coding agent or generating client code:

  • Append .mdx to any docs URL for the raw Markdown, e.g. /docs/quickstart.mdx.
  • /llms.txt — curated discovery index: coverage, why to use it, workflow-oriented doc links (~100 lines).
  • /agents.txt — agent onboarding, auth/pagination rules, and crawl allowlist.
  • /llms-endpoints.txt — every operation with a link to that route's markdown page.
  • /llms-{platform}.txt — per-platform files with parameters, credits, and example curls (from OpenAPI).
  • /llms.json — structured route inventory with parameters, pagination, outcomes, credits, and SDK mappings.
  • /openapi.json — full schema for codegen, validation, and tooling.
  • Capability matrix — dense human-readable appendix of every operation. Prefer /llms.json for agents.

Quickstart

Authenticate with your API key, make a live Social Fetch request, and read the data + meta envelope. Examples in cURL, Node, Python, Go, and Java.