Perplexity
Extract structured data from the Perplexity AI web interface - numbered citations, source snippets, and follow-up questions - as clean JSON.
web_search=false for a faster answer without live citations.Overview
Perplexity is a research-first AI that cites every claim with a numbered source. Use this endpoint to monitor brand mentions in AI search, track which sources Perplexity trusts for your industry, and extract follow-up questions as content and keyword ideas.Unique features
- Numbered citations: Every source is numbered and mapped to inline
[1][2]references in the response text - Source snippets: Full text excerpts from each cited page - not just the URL
- Related questions: Perplexity’s suggested follow-up questions for further research
Example request
Response schema
Includes common response fields plus:Common questions
How do inline citation numbers map to sources?
Theresult field contains inline markers like [1][2]. The number is the 1-based position of the source in the sources[] array - [1] refers to sources[0], [2] to sources[1], and so on.
When should I set web_search=false?
When you need a fast, low-latency answer and don’t need cited sources - for example, generating summaries or using Perplexity as a general-purpose language model. Responses without web search are typically 2–4× faster.
How is Perplexity different from ChatGPT for citation monitoring?
Perplexity returns richer source metadata (full text snippets per source) and uses numbered inline citations that map directly to positions in the response text. ChatGPT returnslinks[] which are sources mentioned but not always position-mapped.Authorizations
Your ScrapeLLM API key passed in the request header.
Query Parameters
The prompt to send to Perplexity. Maximum 4,000 characters.
1 - 4000ISO 3166-1 alpha-2 country code. Routes the request through infrastructure in
that region for localised responses. Defaults to US.
2Enable live web search in Perplexity's response. Set to false for a faster,
non-search answer based on training data only. Defaults to true.
Skip the response cache and always fetch a fresh result.
Maximum seconds to wait for Perplexity to respond. Must be between 10 and 600.
10 <= x <= 600Response
Scrape completed successfully.
Always "perplexity".
"perplexity"
"done" on success.
"done"
Unique identifier for this scrape job.
"job_abc123"
The prompt that was submitted.
"What brands do marketers recommend for email automation?"
The country the request was routed through.
"US"
Plain-text response from Perplexity, with inline citation markers (e.g. [1][2])
corresponding to positions in sources.
"Marketers commonly recommend Mailchimp, HubSpot, and Klaviyo [1][2]..."
Credits consumed by this request. Always 3 for Perplexity.
3
Total end-to-end request duration in milliseconds.
9142.3
true if this response was served from cache.
false
Markdown-formatted response from Perplexity with inline citation markers.
"Marketers commonly recommend **Mailchimp**, **HubSpot**, and **Klaviyo** [1][2]..."
Web sources cited by Perplexity. The position in this array corresponds to the
numbered inline citations in result and result_markdown.
Follow-up questions suggested by Perplexity.
Hosted .md URL for this response. Expires after 24 hours.
"https://scrapellm.com/md/job_abc123.md"