Lesson 433 of 2116
What Perplexity Is: Search-Augmented LLM, Not A Chatbot
Perplexity is built around the idea that every answer should cite its sources. Treating it like ChatGPT misses the point — and the reliability gap that comes with it.
Lesson map
What this lesson covers
Learning path
The main moves in order
- 1Two architectures, one chat interface
- 2search-augmented generation
- 3answer engine
- 4citations
Concept cluster
Terms to connect while reading
Section 1
Two architectures, one chat interface
ChatGPT and Claude generate text from weights and a recent context. Perplexity does that AFTER it has done a live web search and pulled passages into the prompt. The chat box looks identical, but the underlying loop is different — and that difference shows up in every answer.
The Perplexity loop
- 1Take the user query and rewrite it as one or more search queries
- 2Pull the top results, scrape readable text, rank passages by relevance
- 3Hand the passages plus the original query to an LLM with a 'cite your sources' instruction
- 4Render the answer with numbered citations linking back to the sources
- 5Allow follow-up questions that re-query within the same thread
Where the line blurs
ChatGPT and Claude both have search modes now. Google AI Overviews does retrieval too. Perplexity is no longer alone — but it is still the product where retrieval is the default rather than a toggle. That changes how you build a habit around it.
Compare the options
| Tool | Default mode | Citation density | Best when |
|---|---|---|---|
| Perplexity | Search-augmented | Every claim | You want sources up front |
| ChatGPT (search on) | Search-augmented | Selective | Long-form synthesis with web context |
| Claude (web search) | Search-augmented | Selective | Reasoning over search results |
| Default ChatGPT/Claude | Memory-only | None | Brainstorming, writing, code |
Try it: same question, three tools
- Ask Perplexity, ChatGPT, and Claude: 'What changed in EU AI Act enforcement in the last 90 days?'
- Note where each one cites a source vs. asserts from training data
- Click two citations from each tool that uses them. How many actually support the claim?
- Decide which tool you would trust for a memo to a client
Key terms in this lesson
The big idea: Perplexity is not a friendlier ChatGPT. It is a different architecture optimized for sourced answers. Use it when sources matter and treat it like the research librarian, not the brainstorm partner.
End-of-lesson quiz
Check what stuck
15 questions · Score saves to your progress.
Tutor
Curious about “What Perplexity Is: Search-Augmented LLM, Not A Chatbot”?
Ask anything about this lesson. I’ll answer using just what you’re reading — short, friendly, grounded.
Progress saved locally in this browser. Sign in to sync across devices.
Related lessons
Keep going
Creators · 10 min
Spaces: Building Team Knowledge Bases In Perplexity
Spaces are Perplexity's project containers — system prompts, files, and shared chat history. They turn the search engine into a research workspace.
Creators · 10 min
Comet Browser: What It Does That Atlas And Operator Don't
Comet is Perplexity's full browser with a research-native sidebar and an action-capable agent. It plays differently than ChatGPT Atlas or Operator — and the differences matter.
Creators · 8 min
Pages: Turning A Search Into A Sharable Doc
Pages converts a research thread into a publish-ready article with sections, citations, and images. It is content production at the speed of a Perplexity query.