- USDT(TRC-20)
- $0.0
If you've considered trying DeepSeek—the hot new AI app from China—but been put off by concerns around privacy and censorship, you'll be interested to know there's now a tweaked version of the DeepSeek R1 model available inside Perplexity.
Perplexity has long given users a choice of in-house and third-party AIs for answering prompts, and DeepSeek R1 is the latest to make the list. Perplexity cofounder Aravind Srinivas has promised the censorship guardrails have been taken off—so yes, you can ask about Tiananmen Square—and that all associated data (like your prompts and their responses) is stored in the U.S.
DeepSeek is now an option inside Perplexity. Credit: Lifehacker
This is all possible because the DeepSeek models are open-source: Anyone can use and adapt them, free of charge. If you've got the technical know-how, you can modify these AI models yourself, though most of us will want to let the experts (such as those on the Perplexity team) take care of that—if you're interested, there's more on this here.
Specifically, it's the DeepSeek R1 reasoning model that's been added to Perplexity, rather than the more general-purpose V3 model, and both Pro subscribers and those using Perplexity for free can access these models: You get five queries per day for free, or 500 per day if you're paying $20 a month.
DeepSeek cuts off conversations about some topics. Credit: Lifehacker
If you want to see what all the fuss is about when it comes to the speed and performance of DeepSeek, without worrying about your data being piped back to the Chinese government, or finding yourself up against China-centric censorship, this is one of the best ways to go about it. Your data is still held by Perplexity, but in the U.S., and in accordance with the company's privacy policy.
It's also worth noting that other AI providers are following Perplexity's lead: Microsoft is bringing DeepSeek R1 to Windows as part of Copilot, and it's also now been introduced to the You.com AI platform, available alongside other models from the likes of Anthropic, Meta, Grok, and OpenAI.
You can use Perplexity on the web, and on Android and iOS. On the web, your prompts can be inputted in the main text box on the Home tab, or by clicking New Thread on the left. To get an answer from DeepSeek R1, turn the Pro toggle switch on, then pick Reasoning with R1 from the drop-down menu.
Everything works similarly when it comes to the mobile apps. Tap inside the Ask anything box on the main Perplexity tab (the magnifying glass icon, lower left), then tap the Pro label and choose Reasoning with R1. You can then type out and submit your query for the AI.
An answer inside Perplexity produced by DeepSeek R1. Credit: Lifehacker
As DeepSeek R1 is a reasoning model, you'll see that it thinks about your prompts—and explains how it's thinking as it goes. Answers come with website citations attached, where relevant, and at the end of the response you get icons for copying, sharing, and regenerating the answer Perplexity and DeepSeek have given you.
I tested out the Perplexity version of DeepSeek with a few questions that might throw up different answers if they were asked in China, compared to the rest of the world—questions about Taiwan, for example—and there was a free and full flow of information in return. Posing the same prompts on DeepSeek proper returned the following message: "Sorry, that's beyond my current scope. Let’s talk about something else."
You can see DeepSeek R1 'thinking' as it works. Credit: Lifehacker
As for DeepSeek R1's reasoning, it seems fairly robust based on the time I've spent with it—it certainly makes AI responses more trustworthy if you can see how they've been arrived at, and the web sources that have been consulted along the way. You can look all of this up after the answers have been presented, as well as while they're being generated.
We've written before about how well Perplexity stands up to rivals such as Gemini and ChatGPT, blending both generative AI and web search and giving you a choice of different AI models on top. The company has also just launched a Perplexity Assistant feature on Android, so you can use the AI bot in place of Gemini or Bixby.
Full story here:
Perplexity has long given users a choice of in-house and third-party AIs for answering prompts, and DeepSeek R1 is the latest to make the list. Perplexity cofounder Aravind Srinivas has promised the censorship guardrails have been taken off—so yes, you can ask about Tiananmen Square—and that all associated data (like your prompts and their responses) is stored in the U.S.
DeepSeek is now an option inside Perplexity. Credit: Lifehacker
This is all possible because the DeepSeek models are open-source: Anyone can use and adapt them, free of charge. If you've got the technical know-how, you can modify these AI models yourself, though most of us will want to let the experts (such as those on the Perplexity team) take care of that—if you're interested, there's more on this here.
Specifically, it's the DeepSeek R1 reasoning model that's been added to Perplexity, rather than the more general-purpose V3 model, and both Pro subscribers and those using Perplexity for free can access these models: You get five queries per day for free, or 500 per day if you're paying $20 a month.
DeepSeek cuts off conversations about some topics. Credit: Lifehacker
If you want to see what all the fuss is about when it comes to the speed and performance of DeepSeek, without worrying about your data being piped back to the Chinese government, or finding yourself up against China-centric censorship, this is one of the best ways to go about it. Your data is still held by Perplexity, but in the U.S., and in accordance with the company's privacy policy.
It's also worth noting that other AI providers are following Perplexity's lead: Microsoft is bringing DeepSeek R1 to Windows as part of Copilot, and it's also now been introduced to the You.com AI platform, available alongside other models from the likes of Anthropic, Meta, Grok, and OpenAI.
Using DeepSeek inside Perplexity
You can use Perplexity on the web, and on Android and iOS. On the web, your prompts can be inputted in the main text box on the Home tab, or by clicking New Thread on the left. To get an answer from DeepSeek R1, turn the Pro toggle switch on, then pick Reasoning with R1 from the drop-down menu.
Everything works similarly when it comes to the mobile apps. Tap inside the Ask anything box on the main Perplexity tab (the magnifying glass icon, lower left), then tap the Pro label and choose Reasoning with R1. You can then type out and submit your query for the AI.
An answer inside Perplexity produced by DeepSeek R1. Credit: Lifehacker
As DeepSeek R1 is a reasoning model, you'll see that it thinks about your prompts—and explains how it's thinking as it goes. Answers come with website citations attached, where relevant, and at the end of the response you get icons for copying, sharing, and regenerating the answer Perplexity and DeepSeek have given you.
I tested out the Perplexity version of DeepSeek with a few questions that might throw up different answers if they were asked in China, compared to the rest of the world—questions about Taiwan, for example—and there was a free and full flow of information in return. Posing the same prompts on DeepSeek proper returned the following message: "Sorry, that's beyond my current scope. Let’s talk about something else."
You can see DeepSeek R1 'thinking' as it works. Credit: Lifehacker
As for DeepSeek R1's reasoning, it seems fairly robust based on the time I've spent with it—it certainly makes AI responses more trustworthy if you can see how they've been arrived at, and the web sources that have been consulted along the way. You can look all of this up after the answers have been presented, as well as while they're being generated.
We've written before about how well Perplexity stands up to rivals such as Gemini and ChatGPT, blending both generative AI and web search and giving you a choice of different AI models on top. The company has also just launched a Perplexity Assistant feature on Android, so you can use the AI bot in place of Gemini or Bixby.
Full story here:
![[DK]](/data/avatars/s/57/57999.jpg?1720990206){kind=avatar}
