> ## Content Index
> Fetch the complete content index at: https://www.testingcatalog.com/llms.txt
> Use this file to discover other available public pages before exploring further.

# Perplexity AI tests deep research feature, challenging Gemini and ChatGPT
- URL: https://www.testingcatalog.com/perplexity-ai-tests-deep-research-feature-challenging-gemini-and-chatgpt/
- Published: 2025-02-14T11:34:05.000Z
- Updated: 2025-02-14T13:34:27.000Z
- Author: Alexey Shabanov
- Tags: Perplexity News, Latest AI News

Perplexity is preparing to release a deep research feature in Alpha. This feature likely functions similarly to deep research tools on platforms like Gemini or ChatGPT, allowing Perplexity to spend more time analyzing web resources and employing its reasoning capabilities to generate comprehensive reports. It appears that this feature has been temporarily available or limited to a small group of users. Given its Alpha status, it’s possible that the rollout will be gradual, meaning not all users will have immediate access.

> BREAKING 🚨: Perplexity is preparing a release of Deep Research Alpha 🔥  
>  
> As this feature works on Gemini and ChatGPT, perplexity can take more time to go through web resources and utilise its reasoning capabilities to complete the report.  
>  
> It seems like this feature has been… [https://t.co/mUTBAqMkdV](https://t.co/mUTBAqMkdV?ref=testingcatalog.com) [pic.twitter.com/dB3pKxWyWv](https://t.co/dB3pKxWyWv?ref=testingcatalog.com)
> 
> — TestingCatalog News 🗞 (@testingcatalog) [February 14, 2025](https://twitter.com/testingcatalog/status/1890360609318801911?ref%5Fsrc=twsrc%5Etfw&ref=testingcatalog.com)

In a recent video, I tasked Perplexity with categorizing Perplexity-related posts from TestingCatalog. The tool performed commendably. Initially, it sought to understand the task before execution, researching TestingCatalog and recognizing that I publish news about Perplexity. It then visited the website and examined each topic individually. For instance, it analyzed the latest post to determine if a feature had been released and identified the exact announcement date.

Ultimately, Perplexity reviewed numerous posts from the past several months and categorized them as requested. It clarified whether certain features had been released and organized announcements by date. I believe that with more thorough fact-checking post-research and a more explicit initial prompt detailing expectations and research methodology, the results could improve significantly.

Additionally, Perplexity recently introduced an option to export responses into PDF or markdown files, which is particularly beneficial for deep research. For example, users can request a detailed report and then export it into a PDF to share with others.

![Perplexity](https://storage.ghost.io/c/2a/1b/2a1b1782-8506-4d7d-bf53-ad3fb52e2a0f/content/images/2025/02/Screenshot-2025-02-12-at-22.49.12.png)

Export feature

It’s uncertain when this feature will be widely released. However, it presents strong competition to existing offerings from Gemini and ChatGPT. Notably, [Gemini Deep Research](https://www.testingcatalog.com/google-introduces-gemini-deep-research-with-chain-of-thought-reasoning/) is still powered by Gemini 1.5 Pro, which isn’t their latest model, and [Deep Research on ChatGPT](https://www.testingcatalog.com/openai-launches-deep-research-to-automate-multi-step-analysis-in-chatgpt/) is currently only available to Pro subscribers. In this context, Perplexity’s offering provides a more compelling value proposition.

💡

TIP: Get a [10$ discount on Perplexity Pro](https://perplexity.ai/pro?referral%5Fcode=BFE33Y5Y&ref=testingcatalog.com) from TestingCatalog

[Perplexity](https://www.testingcatalog.com/tag/perplexity/) is a conversational search engine that uses large language models to answer queries by sourcing information from the web and providing cited links within its responses. Developed by Perplexity AI, Inc., based in San Francisco, California, the platform aims to deliver accurate, trusted, and real-time answers to user questions.