If you’ve ever toyed with the idea of exploring the latest in artificial intelligence, you might have encountered DeepSeek—a cutting-edge AI app developed in China that has sparked both excitement and apprehension among tech enthusiasts. Concerns over privacy and censorship have long shadowed its use, but a new twist is set to change that narrative. Enter Perplexity, a platform that’s now offering an uncensored version of the DeepSeek R1 model, promising both transparency and an enhanced user experience.

DeepSeek has been making headlines for its robust reasoning capabilities, and the R1 variant is particularly noteworthy. Unlike its predecessor models, DeepSeek R1 is designed to think through prompts, explain its thought process, and back up its responses with reliable citations. This level of introspection is a welcome change in an era where AI answers are often opaque.

What makes this development even more appealing is that the model is now available through Perplexity—a platform known for providing users with multiple AI options, both in-house and from third-party providers. According to Perplexity cofounder Aravind Srinivas, the version of DeepSeek integrated into their service comes with one significant upgrade: the removal of previously embedded censorship guardrails. This means that users can now inquire about historically sensitive topics, such as the events at Tiananmen Square, without the usual filters that might distort or block the response.

The integration of DeepSeek R1 into Perplexity is not just about unfiltered content. It also addresses a major concern for many users: data privacy. In an era where data sovereignty is hotly debated, knowing where your information is stored is crucial. Perplexity assures users that all data—prompts, responses, and any interactions—are stored in the United States. This stands in contrast to earlier apprehensions about Chinese-based AI applications, where there might be fears of data being shared with or controlled by foreign entities.

Perplexity has made it easy for both casual users and tech aficionados to access DeepSeek R1. Whether you’re using the web version or the mobile apps available on Android and iOS, getting started is straightforward. On the web, you can simply input your prompt on the Home tab or click on “New Thread” from the sidebar. To activate the DeepSeek R1 reasoning mode, toggle the Pro switch on and select the corresponding option from a drop-down menu.

For those on the go, the mobile experience is just as user-friendly. A tap on the “Ask anything” box, followed by selecting the Pro label and choosing “Reasoning with R1,” is all it takes to unlock the capabilities of this sophisticated model. The response you’ll receive isn’t just an answer—it’s an explanation of the reasoning behind it, complete with website citations. This transparency allows users to see the “thought process” of the AI, making it easier to verify the information provided.

The accessibility extends beyond casual use. Both free and Pro subscribers have the option to interact with DeepSeek R1, though with different daily query limits (five queries per day for free users and up to 500 per day for Pro subscribers at $20 a month). This tiered approach ensures that the platform caters to a wide range of users, from those curious about AI capabilities to professionals relying on accurate, in-depth reasoning for their work.