perplexity · perplexity · chat & research
Perplexity, with the sources showing.
Answers built from a live web search, with numbered citations you can open, next to the models you'll use once the research is done.
- made by
- Perplexity
- versions
- 5
- from
- 15 tokens
- included on
- Starter+
best at
What Perplexity is good at
Perplexity answers a question by searching the web first and writing second. Every answer arrives with numbered citations, and the numbers in the text link to the pages they came from. That makes it the right tool whenever the honest answer to "how do you know?" needs to be a link.
It is particularly good at questions with a shelf life: what a regulation says this year, how two products compare today, what a company announced last week. It is also a fast way to get oriented in an unfamiliar topic, because the sources double as a reading list.
On LazySusan, Perplexity is usually the first step rather than the whole job. Research a market here, then take the findings to Claude to write the report, or to a picture model for the slide that goes with it, without leaving the workspace.
versions
Perplexity versions on LazySusan
| Version | Use it for |
|---|---|
| Perplexity Auto | Picks per question: Sonar for short ones, Sonar Pro for longer ones, Sonar Reasoning Pro when you ask why, how, or to compare. Deep Research only when you ask for it. |
| Sonar | Fast search for straightforward questions. |
| Sonar Pro | A deeper search for harder questions, with more sources behind the answer. |
| Sonar Reasoning Pro | Search plus step-by-step reasoning, for comparisons and questions that need working out. |
| Sonar Deep Research | Runs many searches and writes a long, cited report. Slow and much more expensive, so it is never chosen for you. |
pricing
What Perplexity costs on LazySusan
| Generation | Tokens |
|---|---|
| Sonar14–18 from a short to a long reply | 15 |
| Sonar Reasoning Pro21–47 from a short to a long reply | 28 |
| Sonar Pro25–74 from a short to a long reply | 38 |
| Sonar Deep Research267–481 from a short to a long reply | 323 |
Costs are for a question of about 300 characters and include the search itself. A Sonar answer is around 15 tokens; a Deep Research report is about 323 tokens, which is why Auto never picks it on its own.
Every plan includes Perplexity. Starter is $8 a month with 1,000 tokens; unused tokens roll over while you're subscribed. Compare plans.
prompts
Prompts to try with Perplexity
// market scan
Who are the main suppliers of compostable coffee pods in the UK right now, and what do they say about certification? Cite each claim.
// regulation
What are the current rules for claiming home-office expenses as a self-employed person in Canada? Link the official sources.
// product research
Compare the three most-reviewed 27-inch monitors for photo editing under $500. Include where each review came from.
// fact check
Is it true that the Great Wall of China is visible from space with the naked eye? Show the sources on both sides.
// deep research
Deep research: the state of grid-scale battery storage in Australia, including major projects, policy and the main open questions.
tips
Getting better results
- 01Ask the question you actually have, with the context that changes the answer (country, year, budget). Perplexity searches better with specifics than with keywords.
- 02Open two or three of the citations before you rely on an answer. Perplexity is good at finding sources, but the source is still what you're trusting.
- 03Words like why, how, explain or compare make Perplexity Auto use Sonar Reasoning Pro, at about 28 tokens a question. For a quick lookup, pick Sonar directly.
- 04Use Deep Research for questions that would otherwise take you an afternoon of tabs. At around 323 tokens a report, it is worth saving for those.
alternatives
When to pick a different tool
also on the api
Perplexity in your own app
Cited, web-grounded answers through the same key and SDK as every other model.
One key, our SDKs for JavaScript and Python, prepaid credits with no subscription. LazySusan API · docs
faq
Perplexity on LazySusan: questions
Yes. Perplexity's Sonar models are included on every LazySusan plan, from Starter at $8 a month. Some features exist only in Perplexity's own app, but the cited search answers are here.
Sonar, Sonar Pro, Sonar Reasoning Pro and Sonar Deep Research, plus Perplexity Auto, which picks between the first three based on your question.
Yes. Answers come with numbered citations, and each number links to the page it came from.
A typical question costs about 15 tokens on Sonar, 28 tokens on Sonar Reasoning Pro and 38 tokens on Sonar Pro. Deep Research is about 323 tokens per report.
Only when you ask for it, for example by writing "deep research" in your question or picking Sonar Deep Research. It is never chosen from the length or wording of a normal question.
Yes. Sonar, Sonar Pro and Sonar Reasoning Pro are in the developer API, so you can add cited web answers to your own app with the same key you use for other models.
get started
Perplexity and every other model, one plan.
One login, one bill and one Library for everything you make, with the right model for each part of the job.
Starter is $8/month. Cancel anytime.