Perplexity subscribers now have access to a new agentic tool.
The company describes Perplexity Computer as a system that integrates all current AI capabilities. It functions as a computer user agent, capable of executing intricate workflows autonomously by utilizing 19 distinct AI models and even generating subagents for specific tasks.
This tool is currently exclusive to Perplexity Max, the company’s premium $200/month subscription tier. Its cloud-based operation may mitigate some security concerns associated with other agentic tools, such as OpenClaw.
While a hands-on demonstration of the tool has not been conducted, example workflows on Perplexity’s website illustrate its ability to manage tasks such as gathering statistical, financial, or legal data, performing analysis, and presenting findings as complete websites or visualizations.
Perplexity held a background briefing with executives to discuss the product and its yearly agenda. A planned demonstration of the tool was canceled due to identified flaws just hours before the event.
This tool signifies an evolution for Perplexity, a company that gained early prominence in the AI sector by integrating advanced models into user-friendly interfaces, notably its search-engine-style answer service. The company later introduced its Comet web browser. An executive noted that competitors, including Google, have since adapted their products to resemble Perplexity’s offerings, viewing this as both an acknowledgment and a competitive challenge.
The company is adapting to a changing industry landscape. After being among the first AI companies to offer advertising, it discontinued this practice late last year, stating that it compromised user trust in the accuracy of its answers. Perplexity’s user base, numbering in the tens of millions, is significantly smaller than OpenAI’s, which reports 800 million weekly users and has started testing ads in ChatGPT this year.
Perplexity executives now indicate a focus on a more specialized user base, developing products for individuals involved in “GDP-moving decisions.” Executives at the briefing, who remained anonymous, emphasized prioritizing enterprise subscriptions, especially for in-depth research.
An executive stated that the company does not discuss Monthly Active Users (MAUs) because its mission is not to acquire the largest possible user base.
Perplexity recently introduced Draco, a new benchmark for complex research tasks, where its deep research offering reportedly outperforms competitors such as Gemini.
Perplexity asserts that it no longer depends on external APIs for its web index, having developed its own AI-optimized search API. The company is intensifying its efforts to present advanced models within a consumer-friendly user experience, contending that coordinating multiple third-party LLMs offers the most cost-effective and precise answers to queries.
A Perplexity executive argued that a “multi-model” approach represents the future, believing that models are specializing rather than becoming commodities. The company observed that its users frequently alternate between models to achieve desired results. For instance, in December 2025, queries for visual outputs were predominantly directed to Gemini Flash, software engineering tasks to Claude Sonnet 4.5, and medical research to GPT-5.1.

Perplexity’s software can automatically select the most suitable LLM for a task, such as choosing one adept at coding over another skilled in drafting marketing copy. Executives also mentioned using Perplexity’s modified open-source Chinese-built LLMs for more cost-effective query responses. This technique, for which the company faced criticism last year for lack of transparency, could be an efficient method for optimizing LLM queries if implemented openly.
The company also provides users with a feature called Model Council, allowing simultaneous queries across multiple models. However, the unit economics of offering unlimited multiple queries at flat subscription rates remain somewhat unclear.
Despite this, executives claim that without significant infrastructure costs and with high margins from user fees, Perplexity expects to maintain competitiveness by allocating tokens to the most appropriate model for each purpose.
Further developments are anticipated: the Perplexity Comet browser is scheduled for release on iOS next month, and the company plans a developers conference, Ask, on March 11 in San Francisco, aimed at promoting third-party API usage.
An executive indicated a shift in focus from daily query counts to recent revenue metrics. Some customers have observed this new emphasis on profitability, with frequent complaints appearing on the Perplexity subreddit regarding new rate limits on both free and subscription product tiers.
However, executives at the briefing dismissed these complaints, with one stating that “any discussions on the free tier being made worse or rate-limited is completely false.”

