Perplexity AI Answer Engine Cheat Sheet

Updated 2026-05-21

Next Topic: Personal Finance and Investing Fundamentals Cheat Sheet

Perplexity AI is a real-time answer engine launched in December 2022 that combines large language models with live web retrieval to deliver synthesized, cited answers instead of a list of links. Unlike traditional search engines that return pages to click through, Perplexity's core philosophy is search → synthesis → verification: every response includes numbered inline citations you can click to confirm. The critical mental model is that Perplexity is a retrieval-augmented system, not a pure reasoning engine — it searches first and generates second, which makes it far better for current facts and source-grounded research than a reasoning-only chatbot, but also means prompt style, focus mode, and source verification habits matter enormously for result quality.

What This Cheat Sheet Covers

This topic spans 20 focused tables and 107 indexed concepts. Below is a complete table-by-table outline of this topic, spanning foundational concepts through advanced details.

Table 1: Core Concept and How It WorksTable 2: Subscription Plans and TiersTable 3: Search Modes and Depth LevelsTable 4: Focus ModesTable 5: Models Available Per QueryTable 6: Spaces and ThreadsTable 7: Perplexity PagesTable 8: Perplexity LabsTable 9: File Upload and Document Q&ATable 10: Perplexity ShoppingTable 11: Perplexity Comet BrowserTable 12: Perplexity ComputerTable 13: Sonar API for DevelopersTable 14: Discover Feed and Trending TopicsTable 15: Privacy, Data Controls, and the Ad ModelTable 16: Comparison with CompetitorsTable 17: Prompt Patterns and Research TechniquesTable 18: Citation Workflows and Source VerificationTable 19: Common Pitfalls and GotchasTable 20: Daily Use Cases

Table 1: Core Concept and How It Works

Perplexity's underlying architecture sets it apart from both traditional search engines and pure AI chatbots — understanding the pipeline demystifies why some queries shine and others need follow-up.

Concept	Example	Description
Answer Engine	Ask "What is CRISPR gene editing?" → get a 3-paragraph cited answer, not 10 blue links	• Perplexity's self-description • it synthesizes a direct answer from multiple sources rather than returning a ranked list of pages to browse
Retrieval-Augmented Generation (RAG)	User query → live web search → retrieved pages fed to LLM → cited answer	The core pipeline: real-time retrieval grounds the LLM's output in current sources, reducing hallucination and keeping answers up to date.
Inline Citations	Numbered superscripts `[1][2][3]` inside every answer linking to source URLs	• Every factual claim is linked to the web page it came from • users can click any citation to verify the original source
Real-Time Web Search	Asking about today's stock price or breaking news returns current data, not stale training cutoff results	Perplexity fetches pages live at query time, unlike base LLMs whose knowledge is frozen at training.

Table 1: Core Concept and How It Works

Concept	Example	Description
Answer Engine	Ask "What is CRISPR gene editing?" → get a 3-paragraph cited answer, not 10 blue links	• Perplexity's self-description • it synthesizes a direct answer from multiple sources rather than returning a ranked list of pages to browse
Retrieval-Augmented Generation (RAG)	User query → live web search → retrieved pages fed to LLM → cited answer	The core pipeline: real-time retrieval grounds the LLM's output in current sources, reducing hallucination and keeping answers up to date.
Inline Citations	Numbered superscripts `[1][2][3]` inside every answer linking to source URLs	• Every factual claim is linked to the web page it came from • users can click any citation to verify the original source
Real-Time Web Search	Asking about today's stock price or breaking news returns current data, not stale training cutoff results	Perplexity fetches pages live at query time, unlike base LLMs whose knowledge is frozen at training.