SocialQuery Prism Scraper API

32 Composite intelligence endpoints for every platform. Prism turns raw endpoints into finished social intelligence: review integrity, AI visibility, share of voice, creator vetting, launch tracking, and more — all as API calls.

Get your API key Read the docs

Development preview. Composite recipes are not yet available. The illustrations below show the intended response structure, not live results.

cURLPythonTypeScript

curl https://socialquery.dev/v1/prism/lookup \
  -G --data-urlencode "url=https://www.youtube.com/watch?v=A9TikdsD5eg" \
  -H "x-api-key: $SOCIALQUERY_API_KEY"

{
  "resolved": { "platform": "youtube", "endpoint": "/v1/youtube/video" },
  "legs": [{ "status": "ok", "provider_route": ["youtube"] }],
  "data": { "provider": "youtube" }
}

The problem today

Social data is fragmented. Right now, you are the integration layer.

Every platform speaks a different dialect: its own auth, shape, paging rules, failure modes, and maintenance schedule.

auth x N

A new key, header, and provider behavior for every source you add.

schema x N

A dozen response shapes to normalize before one answer exists.

maintenance x N

Each source can change independently and break a workflow overnight.

How Prism works

One call. Many platforms. One enriched answer.

One request enters the composite engine. Prism resolves entities in parallel, runs dependent calls, joins identity across platforms, and folds the output into one billed response.

Your single call

GET /v1/prism/creator-card?handle=@x

Composite Engine

Fan out, fuse, score, and validate

Wave 1

Profiles, posts, comments, search, reviews

Wave 2

Dependent calls and identity joins

Unified Result

One object with per-leg credits

Under the hood

One call. Every source. Named in the response.

You hit one endpoint. Prism queries every relevant source, fuses and dedupes the results, scores them with deterministic code, and returns a per-source ledger you can audit.

Review integrity

/v1/prism/review-integrity

Amazon, Trustpilot, forums

A-F grade

Share of voice

/v1/prism/share-of-voice

X, Reddit, YouTube, TikTok

conversation split

AI visibility

/v1/prism/ai-visibility

ChatGPT, Gemini, Perplexity

visibility score

Crisis radar

/v1/prism/crisis-radar

social, forums, news

rising risk

Brand mentions

/v1/prism/brand-mentions

all relevant sources

ranked feed

Forum search

/v1/search/forums

Reddit, HN, communities

evidence list

Source ledger

reddit: ok

youtube: ok

x: retry

tiktok: ok

refund: none

The moat

A shared design for multi-source workflows.

Single-shot threads

Return the thread, comments, and community context without assembling them client-side.

Receipts by source

Every source leg is named so teams can audit what contributed to the answer.

Deterministic scoring

The final grade, percentage, or ranking is computed in code, not guessed by copy.

What you can build

Flagship recipes. One call, one finished job.

Every recipe returns a grade, percentage, ranked feed, or finished object instead of raw rows to assemble.

Detect fake reviews

Return an integrity grade across commerce, review sites, and forums.

/v1/prism/review-integrity

Track AI visibility

Measure how answer engines describe your brand and competitors.

/v1/prism/ai-visibility

Measure share of voice

Compare the conversation you own against competing brands.

/v1/prism/share-of-voice

Monitor brand mentions

Follow volume, sentiment, and source movement from one call.

/v1/prism/brand-mentions

Search the forums

Find the discussions shaping buyer intent.

/v1/search/forums

One URL, full data

Resolve a source into the surrounding context your product needs.

/v1/prism/lookup

AI answers with receipts

Return a synthesized answer with source evidence attached.

/v1/prism/answers

Crisis radar

Spot fast-moving negative signals before they become a report.

/v1/prism/crisis-radar

Who builds on Prism