SocialCrawl API Documentation

SocialCrawl returns public data from 65 platforms through 575 endpoints. Every request uses one x-api-key header, and every response comes back in the same JSON shape, so one parser covers every platform.

You do not need an account to try it. Press Run below. If you are signed in, your own API key is already filled into every sample on this page.

Make your first call

curl "https://www.socialcrawl.dev/v1/tiktok/profile?handle=charlidamelio" \  -H "x-api-key: YOUR_API_KEY"

Three things just happened. You sent one GET request with one header. The platform answered in well under a second on a normal connection. The call cost 1 credit, and your balance came back in the same response, so you always know where you stand.

Same call, different platform

Change one path segment and the same request reads a different platform. Same header, same handle parameter, same response shape.

curl "https://www.socialcrawl.dev/v1/tiktok/profile?handle=nasa" \
  -H "x-api-key: YOUR_API_KEY"

curl "https://www.socialcrawl.dev/v1/instagram/profile?handle=nasa" \
  -H "x-api-key: YOUR_API_KEY"

curl "https://www.socialcrawl.dev/v1/youtube/channel?handle=nasa" \
  -H "x-api-key: YOUR_API_KEY"

curl "https://www.socialcrawl.dev/v1/twitter/profile?handle=nasa" \
  -H "x-api-key: YOUR_API_KEY"

You do not learn four APIs. Field coverage differs where the platforms differ, and each endpoint's exact response shape is one free call away (see below).

Read the response

Every response, success or failure, has the same top level. The platform data is always under data.

{
  "success": true,
  "platform": "tiktok",
  "endpoint": "/v1/tiktok/profile",
  "data": {
    "author": { "username": "nasa", "followers": 3900000 }
  },
  "credits_used": 1,
  "credits_remaining": 99,
  "request_id": "req-abc123",
  "cached": false
}

List endpoints add a pagination block. Send next_cursor back as cursor and stop when has_more is false. That rule is the same on every list endpoint. See Pagination and Response schema.

When an endpoint supports a computed field and the required source inputs are present, the response also carries it under computed, for example engagement_rate. A field that cannot be computed is null, never a guess. See Computed fields.

Explore the catalogue for free

Every /v1/utility/* endpoint and /v1/credits/balance cost 0 credits. They answer in about 150 ms and are generated from the endpoint registry at request time, so they cannot drift from what is callable. Use them instead of reading 575 reference pages.

# Everything you need for a first call, in one JSON document
curl "https://www.socialcrawl.dev/v1/utility/quickstart?platform=instagram" \
  -H "x-api-key: YOUR_API_KEY"

# Search every endpoint by keyword. Try "comments", "reviews", "transcript"
curl "https://www.socialcrawl.dev/v1/utility/endpoints?search=comments" \
  -H "x-api-key: YOUR_API_KEY"

# The full guide for one endpoint: parameters, cost, paging, an example call
curl "https://www.socialcrawl.dev/v1/utility/endpoint?id=youtube/channel" \
  -H "x-api-key: YOUR_API_KEY"

# Your balance
curl "https://www.socialcrawl.dev/v1/credits/balance" \
  -H "x-api-key: YOUR_API_KEY"

Hand it to your AI agent

If Claude, Cursor, ChatGPT, or another agent writes your code, give it the same catalogue. The MCP server and the Agent Skill both call it before every request, so the agent never guesses a path or a parameter. The Using an AI agent page has a prompt you can paste as it is.

# Claude Code (hosted MCP, nothing to install)
claude mcp add --transport http socialcrawl https://mcp.socialcrawl.dev/mcp \
  --header "Authorization: Bearer YOUR_API_KEY"

# Any MCP client, local package
npx -y socialcrawl-mcp

Every docs page is also plain Markdown: append .md to its URL. The machine index is at /llms.txt, the per-platform context is at /v1/utility/llms?platform=tiktok, and the full spec is at /v1/openapi.json. See AI agents and Skills and MCP.

What a call costs

New accounts start with 100 credits and no card. Most endpoints cost 1 credit. Cache hits cost 0. Failed calls and empty results are refunded automatically. Invalid parameters are rejected before billing.

A small set of endpoints has a flat or metered price instead of a tier. The exact figure for every endpoint is in Endpoint pricing and in the credits field of /v1/utility/endpoints. Rate limits are 600 requests per minute and 50 in flight per key. See Credits and Rate limits.

When something goes wrong

Errors use the same JSON shape with success: false. Branch on error.type, never on the message text. Every error carries a doc_url that links to the fix.

ENDPOINT_NOT_FOUND means we do not have it. RESOURCE_NOT_FOUND means the platform does not have it. The full table, with which codes are safe to retry, is in Errors.

Next

You made a call. The Quickstart walks you through your first 100 calls: your own key, several platforms in one script, one search across every platform, and the free catalogue from your terminal.

Which platforms does SocialCrawl support?

Platform	Endpoints
LinkedIn	45	View →
Instagram	37	View →
TikTok	36	View →
Prism	33	View →
YouTube	29	View →
Facebook	24	View →
Web Scraping	22	View →
US Congress Trades	19	View →
Klarna	18	View →
Tripadvisor	16	View →
Twitter/X	15	View →
Naver	14	View →
Reddit	14	View →
GitHub	12	View →
Gumtree	11	View →
Jobs	11	View →
Sephora	11	View →
Content Analysis	10	View →
Google	10	View →
AliExpress	9	View →
Apple App Store	9	View →
Google Play	9	View →
Amazon	8	View →
Douyin	8	View →
Finance	7	View →
G2	7	View →
Quora	7	View →
H&M	6	View →
Spotify	6	View →
Threads	6	View →
Google Shopping	5	View →
Kohl's	5	View →
Pinterest	5	View →
Rumble	5	View →
Target	5	View →
TikTok Shop	5	View →
Walmart	5	View →
Yelp	5	View →
Apple Music	4	View →
Etsy	4	View →
Hacker News	4	View →
Home Depot	4	View →
Tavily	4	View →
Twitch	4	View →
Utility	4	View →
Bluesky	3	View →
Google Trends	3	View →
Kwai	3	View →
Telegram	3	View →
Truth Social	3	View →
Wayfair	3	View →
eBay	2	View →
Snapchat	2	View →
Trustpilot	2	View →
Google News	1	View →
Kick	1	View →
Komi	1	View →
LinkBio	1	View →
LinkMe	1	View →
Linktree	1	View →
On-Page	1	View →
Perplexity	1	View →
Pillar	1	View →
Polymarket	1	View →

Frequently asked questions

How do I authenticate with the SocialCrawl API?

Pass your API key in the x-api-key header on every request. Sign up at socialcrawl.dev, open Dashboard → API Keys, and create a key. There is no OAuth flow, token exchange, or session to manage.

Which platforms does the SocialCrawl API support?

SocialCrawl covers 65 platforms through 575 endpoints (including TikTok, Instagram, YouTube, Facebook, Twitter/X, LinkedIn, Reddit, Threads, Pinterest, and Twitch) plus research sources like GitHub, Hacker News, Tavily, and Perplexity.

What does the SocialCrawl response envelope look like?

Every response (success or error) uses the same JSON envelope: success, platform, endpoint, data, credits_used, credits_remaining, request_id, and cached. List responses return data as { items, next_cursor }, and computed fields like engagement_rate arrive pre-calculated on every author and post.

How do I use SocialCrawl with Claude, Cursor, or other MCP clients?

Install the MCP server with npx -y socialcrawl-mcp. In Claude Code, run: claude mcp add --scope user socialcrawl -- npx -y socialcrawl-mcp. In Claude Desktop, Cursor, VS Code, or Windsurf, add socialcrawl-mcp to your mcpServers config with your SOCIALCRAWL_API_KEY.

How many free credits do I get when I sign up?

Every new account receives 100 welcome credits automatically on signup. Standard endpoints cost 1 credit per request, so that is up to 100 free requests to evaluate the API, and cache hits cost 0 credits.