Perplexity AI Releases A New "Pro Search" Feature
Perplexity just rolled out a new and improved Pro Search feature with improved problem-solving capability through multi-step reasoning and a more advanced math and programming capability.
Perplexity AI released some huge upgrades and enhancements to Perplexity Pro Search. The update includes an improved problem-solving capability through multi-step reasoning, a more advanced math and programming capability, and more in-depth and comprehensive search results.
I would assume you already know what Perplexity is and what it is for, but in case you don’t know it yet, here’s a quick introduction.
What is Perplexity?
Perplexity is an AI-powered search and question-answering platform that aims to provide accurate, up-to-date information by combining elements of traditional search engines with natural language AI capabilities.
Perplexity AI was launched in 2022 with the aim to compete with Alphabet Inc.’s Google Search. The search capability is based on OpenAI’s GPT-3.5 model and a standalone LLM with NLP capabilities for its free product. The pro version accesses GPT-4o, Claude 3.5 Sonnet, and others.

I have been using Perplexity since last year and honestly, it’s a really good tool if you want to have straightforward answers without the clutter of sponsored posts or the need to sift through links to find the desired answer.
What’s new in Perplexity Pro?
Today, Perplexity rolled out some updates and enhancements to its Pro Search feature.
Multi-step reasoning
Advanced math and programming capabilities
Quick Search vs Pro Search
Let’s get into the details of these features.
Multi-step Reasoning
Pro Search is now more powerful at solving complex problems through multi-step reasoning. It can understand when a question needs careful planning, break down the problem into smaller goals, and deliver thorough answers efficiently.
Take a look at this example:
Prompt: The rise and fall of Stability AI