When you purchase through links on our site, we may earn an affiliate commission.Heres how it works.
Mobile: Perplexity offers aniPhone app, compatible with iPhone and iPad devices running iOS 13.0 or later.
The app supports multiple languages, including English, French, German, Japanese, Korean, and Spanish.
Users on Android can download the Perplexity app from theGoogle Play Store.
The AI Phone is anticipated to launch in the second half of 2025.
The answers are phrased in a conversational way, similar to interacting with a knowledgeable assistant.
This makes searching more conversational and intuitive.
The inline citations support Perplexitys answers.
These appear as clickable footnotes that lead to the original web content.
The interface is useful in that it supports follow-up queries in a thread that encourages conversational search.
After a you get an answer, you could follow up with a related question or clarify something.
Focused searches can be made on specific sources or domains.
AndPerplexity’s deep research featuresgive you even more of a deep dive into just about any subject.
This versatility helps tailor results to the nature of the query e.g.
researching scholarly information vs. getting public opinions.
Creating a free account lets you search your chat history and past conversations.
This is useful if you want to revisit previous answers or build on earlier queries.
Pro users can access more powerful AI models and extra tools.
Pro users also get higher usage limits and faster responses.
and ask questions about their content, blending internal data with web search.
This is powerful for analyzing reports or datasets alongside online information.
Beyond chat and search
Perplexity isnt just limited to queries.
The model also has the ability to generate content in various forms.
To enhance usefulness, Perplexity has rolled out specialized search hubs.
Limitations and concerns
Despite its impressive capabilities, Perplexity AI is not without its limitations and criticisms.
These AI mistakes are often calledhallucinations.
Because the model is synthesizing information, it might occasionally misquote a source or combine facts incorrectly.
Perplexity tries to pull from “trustworthy” web sources, but determining trustworthiness is tricky.
If the underlying web content on a topic is biased or skewed, the answer will reflect that.
The AI model itself also has biases learned from training data.
This can show up in subtle ways, like how questions about controversial topics are answered.
Perplexity logs queries to improve the service (as most search engines do).
Perplexitys quality is tied to what it can find on the web.
Perhaps the most publicized concern has been from news publishers and content creators.
Perplexity has been accused of scraping content without proper permission or attribution.
The emergence of Perplexity has put pressure on established players to innovate.
Perplexity will likely integrate newer and more powerful AI models as they become available.
We may also see improved handling of multimedia queries.
