DeepSeek R1, a new reasoning model from a Chinese startup, is available on Perplexity AI, enhancing deep web research with lower costs and robust privacy protection. The model claims to perform on par with industry leaders while ensuring no data is sent to China. The launch reflects rising competition in the AI space, with user security in mind.
DeepSeek R1, the cutting-edge reasoning model from a Chinese AI startup, has made its debut on the US-based search engine Perplexity AI. Offering superior performance akin to industry leaders at a fraction of the cost, it now enhances deep web research. This model features a new Pro Search reasoning mode selector, enhancing user experience with transparency in its reasoning process.
The company announced, “We’re increasing the number of daily uses for both free and paid plans as we add more capacity throughout the day. Stay tuned for updates.” They emphasize that DeepSeek operates out of US and EU data centers, ensuring data privacy with claims like, “Your data never leaves Western servers. Your privacy and data security is our priority.”
Aravind Srinivas, CEO of Perplexity AI, recently acknowledged DeepSeek AI’s rise to the 1 spot on the App Store, expressing excitement over the competitive landscape against established AI models like ChatGPT. He indicated great anticipation for utilizing DeepSeek’s models this year, promoting a sense of innovation and collaboration in the AI field.
John Srinivas addressed user concerns regarding data security, asserting, “All DeepSeek usage in Perplexity is through models hosted in data centers in the USA and Europe. DeepSeek is open-source. None of your data goes to China.” This focus on transparency reassures users about the safety of their data.
Launched by a small startup in Hangzhou just last week, DeepSeek R1 promises exceptional affordability, being 20 to 50 times cheaper than its counterparts like OpenAI’s models. Leading benchmarks showcase its prowess, reporting parity or superiority over rivals across various assessments, including AIME 2024 for mathematical problems and MMLU for general knowledge queries.
The development of DeepSeek R1 stems from the growing need for efficient AI models that not only perform excellently but also prioritize user privacy. As competition in the AI space intensifies, particularly against giants like OpenAI, startups are compelled to innovate rapidly. DeepSeek’s introduction on Perplexity seeks to bridge the gap between high performance and affordability while maintaining strict data privacy standards for users outside of China.
In summary, DeepSeek R1 presents an innovative solution in the AI marketplace by providing comparable services to leading models at a significantly lower cost. With enhancements aimed at improving user experience and a robust commitment to data privacy, it sets a new standard for how AI can support deep web research. As the AI landscape evolves, DeepSeek’s unique approach could redefine user expectations and opportunities in the field.
Original Source: www.livemint.com
