Klu raises $1.7M to empower AI Teams  

What is Perplexity.ai?

by Stephen M. Walker II, Co-Founder / CEO

What is Perplexity.ai?

Perplexity AI is a technology company that provides a platform for information discovery and sharing. It leverages the power of artificial intelligence (AI) to deliver credible, up-to-date answers to users' queries.

The company offers a mobile application, available on both Android and iOS platforms, that allows users to ask anything and get straight to credible, up-to-date answers. The app syncs across devices and uses AI technologies like OpenAI's GPT-4 and Anthropic's Claude 2. It features a Perplexity Copilot for guided AI search, voice or text input, thread follow-up for deeper understanding, cited sources for every answer, and a library for curating discoveries.

In addition to the mobile app, Perplexity AI also offers a browser extension called Perplexity - AI Companion. This tool provides instant page summaries, quick queries, contextual understanding, shareable insights, and dynamic conversations. It's designed to serve users' curiosity instantly and has been praised for its usefulness in both work and personal contexts.

Perplexity AI has recently raised significant funding, with a valuation of $520 million. The funding round included investors such as Nvidia and Amazon founder Jeff Bezos. The company plans to use the funding to aid in customer adoption and global expansion. In 2023, Perplexity AI served more than half a billion queries, demonstrating its growing influence in the search and AI sectors.

What does Perplexity mean?

Perplexity quantifies the predictive accuracy of language models in natural language processing and machine learning. It is the inverse of the geometric mean of the model's probability distribution for all possible outcomes, given an input. A lower perplexity indicates a model's higher predictive accuracy, with a score of 1 denoting perfect prediction. Conversely, higher scores reflect reduced model performance and greater unpredictability.

What are its key features?

Perplexity AI offers a range of key features designed to provide users with accurate, comprehensive, and contextually relevant information. Here are the main features:

  1. AI-Based Search Engine — Perplexity AI uses state-of-the-art natural language processing (NLP) and machine learning algorithms to understand and analyze the context of search queries.

  2. Contextual Understanding — The platform generates useful and contextually accurate responses by considering the entire conversation history. It selects the best answer from a range of sources and presents a concise summary.

  3. Efficient Information Gathering — Unlike traditional search engines, Perplexity AI allows users to ask direct questions and receive clear, accurate answers backed by curated sources.

  4. Comprehensive and Transparent Answers — Perplexity AI provides complete answers and offers transparency about the sources, allowing users to trace the origin of the information.

  5. Personalized User Experience — The platform offers a personalized user experience, tailoring its responses to the user's specific needs and preferences.

  6. Chatbot-Style Search Engine — Perplexity AI operates as a chatbot-style search engine, providing a conversational interface for users to interact with.

  7. Information Discovery and Sharing — The platform unlocks the power of knowledge with information discovery and sharing, providing users with a wealth of information at their fingertips.

  8. Mobile App and Browser Extension — Perplexity AI is accessible through a mobile application and a browser extension, offering users flexibility and convenience in accessing information.

  9. Problem-Solving Abilities — The platform is designed to assist various professions like researchers, writers, artists, musicians, and programmers in multiple tasks such as answering questions, generating text, writing creative content, and summarizing text.

  10. Real-Time Data Fetching — Perplexity AI has real-time data fetching capability, making it a reliable tool for current information needs.

What are the sources that perplexity ai uses to generate information?

Perplexity AI generates information from a wide array of sources across the web. These include major platforms like Wolfram Alpha, Wikipedia, academic literature and research papers, YouTube, Reddit, and more. The platform is designed to provide up-to-date, credible information, and it cites the sources for every piece of information it generates.

In addition to these sources, Perplexity AI leverages the power of large language models such as OpenAI's GPT-4 and Claude 2 to generate human-like text and provide direct responses to user queries.

It's important to note that while Perplexity AI strives for accuracy and provides source links for its information, users are still advised to conduct their own fact-checks. The tool can occasionally use outdated sources or provide generalized answers, making it essential for users to validate the information themselves.

How does it work?

Perplexity AI operates as a conversational AI platform that utilizes multiple large language models (LLMs) to provide users with instant, reliable answers to their queries. It integrates models from OpenAI, such as GPT-4, and Anthropic's Claude 2, among others, to generate responses that are accompanied by sources and citations for transparency and trust.

The platform is accessible through a mobile application available on both Android and iOS, which features a Perplexity Copilot for guided AI search, voice or text input, thread follow-up for deeper understanding, and a library for curating discoveries. Additionally, Perplexity AI offers a browser extension called Perplexity - AI Companion, which provides instant page summaries, quick queries, contextual understanding, and dynamic conversations directly from the user's browser.

Perplexity AI's approach to search is conversational, allowing users to ask follow-up questions to refine their search and get more relevant answers. This is in contrast to traditional search engines that may provide a list of links without direct answers. The platform's ability to use multiple LLMs enables users to switch between different models, potentially offering a broader range of perspectives and responses.

What are its strengths and limitations?

The platform's conversational approach to search, providing direct answers to queries with curated sources, saves users time by delivering precise and concise information. However, Perplexity AI has its limitations. It may not have as many features as some AI tools, and like all AI, it can produce incorrect answers or exhibit bias. Responses can be inconsistent or overly brief, and it may struggle with complex scientific problem-solving or subjective questions. Additionally, its performance can vary with different test sets, and there are rate limits on model usage that could restrict the number of allowable requests within a given timeframe.

What is PPLX Labs API?

The pplx-api is a fast and efficient API developed by Perplexity Labs for open-source Large Language Models (LLMs). It is designed to be easy to use, requiring no deep knowledge of C++/CUDA or access to GPUs, while still providing state-of-the-art performance. The API can handle a daily load of over one million requests, processing almost one billion tokens daily. It is used to power Perplexity Labs, a model playground serving various open-source models. In the future, pplx-api will support custom Perplexity LLMs and other open-source LLMs, custom Perplexity embeddings and open-source embeddings, and will have a dedicated API pricing structure.

The pplx-api is also used in various applications, such as in LlamaIndex, where it is used to create a simple interface for different functionalities. It is also used in Google Collab for AI personalization.

Please note that there is another pplx namespace in the C++ REST SDK, a Microsoft project for cloud-based client-server communication in native code using a modern asynchronous C++ API design. This pplx namespace provides classes and functions that give you access to the Concurrency Runtime, a concurrent programming framework for C++. However, this is not related to the pplx-api developed by Perplexity Labs.

What are PPLX Labs API features?

Perplexity Labs, offer a platform for users to test and experiment with large language models (LLMs) for free. Perplexity Labs offers two online models, pplx-7b-online and pplx-70b-online, which are focused on delivering helpful, up-to-date, and factual responses. As of January 2024, the playground supports:

  • pplx-7b-online
  • pplx-70b-online
  • pplx-7b-chat
  • pplx-70b-chat
  • mistral-7b-instruct
  • codellama-34b-instruct
  • llama-2-70b-chat
  • llava-v1.5-7b-wrapper
  • mixtral-8x7b
  • mistral-medium

The pplx-api developed by Perplexity Labs has several key features that make it a powerful tool for developers and businesses:

  1. High Performance — The pplx-api is designed for high throughput and low latency. It offers up to 2.9x lower latency than Replicate and 3.1x lower than Anyscale. It is capable of serving 1 billion tokens in a production environment daily.

  2. Ease of Use — The API is user-friendly and does not require deep knowledge of C++/CUDA or access to GPUs. Developers can use state-of-the-art open-source models off-the-shelf and get started within minutes.

  3. Support for Multiple Models — The pplx-api currently supports various models including Mistral 7B, Llama2 13B, Code Llama 34B, and Llama2 70B. It is also OpenAI client-compatible for easy integration with existing applications.

  4. Future Enhancements — In the near future, the pplx-api will support custom Perplexity LLMs and other open-source LLMs, custom Perplexity embeddings and open-source embeddings, and will have a dedicated API pricing structure.

  5. Secure Access — The API can be accessed using HTTPS requests. Users can generate an API key through the Perplexity Account Settings Page and send it as a bearer token in the Authorization header with each pplx-api request.

  6. Integration with Other Platforms — The pplx-api can be integrated with other platforms like Google Collab for AI personalization and LlamaIndex for creating a simple interface for different functionalities.

  7. SDKs for Easy Integration — There are SDKs available for easy integration of the pplx-api into your projects. For example, there is a PHP SDK that allows seamless integration of advanced AI-powered features into your PHP projects.

These features make the pplx-api a robust and efficient tool for developers and businesses looking to leverage the power of large language models.

A Deep Dive into Perplexity AI

Perplexity AI is an AI-driven search tool that leverages OpenAI's language models to provide precise, relevant, and reliable answers to user queries. Developed by Aravind Srinivas, with experience from Google, OpenAI, and DeepMind, Perplexity AI aims to enhance the online search experience by offering conversational interactions with a knowledgeable assistant.

How Perplexity AI Works

Perplexity AI utilizes the GPT-3.5 language model to understand and respond to queries in natural language. It indexes internet content to deliver contextually relevant answers, allowing users to select source types for trustworthy information. The platform integrates web search with conversational AI, distinguishing it from traditional search engines and other AI platforms like ChatGPT.

Key Features and Benefits

Perplexity AI offers a chatbot-style search engine interface, advanced AI technology using GPT-3.5, and natural language understanding (NLU) capabilities. It provides concise answers with source links, supports multiple AI models for Pro accounts, and maintains an ad-free user interface. These features cater to a seamless and focused search experience.

Use Cases and Applications

The platform serves students, researchers, journalists, and marketers, assisting with academic research, fact-checking, content validation, and market research. Perplexity AI's integration into apps like news summarizers and educational tools demonstrates its versatility and ability to enhance workflows across various domains.

Plans and Pricing

Perplexity AI offers a free core search experience and Pro Perplexity accounts with advanced features starting at $20 per month. The platform provides a free trial for Pro accounts, allowing users to test advanced models and functionalities without immediate commitment.

Support and Community

Perplexity AI ensures a comprehensive support system with extensive online resources, FAQs, customer service contact, and community forums. The platform's responsiveness and depth of information available are highlighted in user reviews, emphasizing its commitment to user satisfaction.

Integrations and API

The platform offers a Chrome extension for browser integration, email client summarization, and news app condensation. Perplexity AI's API enables developers to create custom search solutions, enhance data retrieval, and streamline workflows, showcasing the platform's potential for innovation.

Customer Feedback

Users praise Perplexity AI for its effectiveness, ease of use, and reliability, particularly noting the inclusion of source links and the option to switch between AI models. While some users suggest improvements for complex queries, the overall sentiment is positive, with the platform seen as a robust tool for various professional and educational needs.

Frequently Asked Questions

Perplexity AI differentiates itself by focusing on accurate, sourced information using OpenAI's GPT-4 model, as well as several of their own custom LLMs built on Llama 2 70B. It prioritizes user privacy and data protection, ensuring a secure search experience. The platform's source citing mechanism and real-time search capabilities complement traditional search engines, offering a more interactive and precise experience.

Useful Links and Resources

Users can access a range of resources, including the official website, pricing page, Chrome extension, app download pages, user guides, and API documentation. These resources support users in maximizing the platform's capabilities for information retrieval and verification.

The Future of AI-Powered Search

Perplexity AI represents the future of AI-powered search, with potential advancements in multimodal data integration, contextual and predictive search, and customization across sectors. The platform's commitment to data privacy, ethical AI usage, and accessibility ensures it remains a valuable tool for augmenting human intelligence and fostering equitable access to information.

More terms

What is a bag-of-words model?

A bag-of-words model is a simple way to represent text data. It is a representation where each word in the text is represented by a number. The order of the words is not taken into account, so this model is also called a bag-of-words model.

Read more

What is Bayesian programming?

Bayesian programming is a formalism and methodology used to specify probabilistic models and solve problems when less than the necessary information is available. It is a statistical method to construct probability models and solve open-ended problems with incomplete information. The goal of Bayesian programming is to express human intuition in algebraic form and develop more intelligent AI systems.

Read more

It's time to build

Collaborate with your team on reliable Generative AI features.
Want expert guidance? Book a 1:1 onboarding session from your dashboard.

Start for free