Perplexity’s new Computer is another bet that users need many AI models

Published: (February 27, 2026 at 12:00 PM EST)
4 min read
Source: TechCrunch

Source: TechCrunch

New Perplexity Computer

Perplexity Computer “unifies every current AI capability into a single system.” It is a cloud‑based computer‑user agent that can execute complex workflows independently using 19 different AI models, even creating sub‑agents to handle specific problems.

Availability and Pricing

The tool is available now only on the company’s highest subscription tier, Perplexity Max, which costs $200 / month. Because it runs entirely in the cloud, it may avoid some of the security concerns that affect other agentic tools such as OpenClaw.

Capabilities and Example Workflows

Perplexity’s website showcases example workflows that involve:

  • Collecting statistics, financial or legal data
  • Creating analysis and visualizations
  • Publishing findings as finished websites or interactive dashboards

You can explore the examples here: example workflows.

Company Context and Strategy

Perplexity first gained attention by wrapping frontier models in familiar user interfaces, most notably its search‑engine‑like answer service. It later launched the Comet web browser (summer 2023). Competitors such as Google have begun adapting their products to resemble Perplexity’s approach, which the company sees as both a compliment and a threat.

The firm is also adjusting its business model. After abandoning its advertising program late last year—citing concerns that ads eroded trust in answer accuracy—Perplexity now focuses on a boutique set of users who make “GDP‑moving decisions.” Its user base runs in the tens of millions, far smaller than OpenAI’s reported 800 million weekly users.

Executive Briefing Insights

In a background briefing (Boston, MA, June 9 2026), executives—who asked to remain unnamed—emphasized:

  • Prioritizing enterprise subscriptions for deep research.
  • Not chasing monthly active users (MAUs): “You don’t hear us talk about MAUs ever, because we’re not actually on a mission to get as many users as possible.”

They also highlighted the recently released Draco benchmark for complex research tasks, where Perplexity’s own deep‑research offering outperformed competitors such as Gemini.

Multi‑Model Strategy

Perplexity no longer relies on external APIs for its web index; it now operates its own AI‑optimized search API. The company argues that orchestrating multiple third‑party LLMs yields the most cost‑effective and accurate answers.

“Multi‑model is the future,” said a Perplexity executive. “Models are specializing, not commoditizing.”

Internal data (December 2025) shows distinct model preferences:

Use casePreferred model
Visual outputsGemini Flash
Software engineeringClaude Sonnet 4.5
Medical researchGPT‑5.1

Visualization of model usage by Perplexity users over time. Image Credits: Perplexity

If one LLM excels at coding and another at drafting marketing copy, Perplexity’s software can automatically select the ideal model. The company also runs modified open‑source Chinese‑built LLMs to answer queries more cheaply—a practice that drew criticism last year for lack of transparency, though Perplexity now claims to be open about the approach.

Model Council

A feature called Model Council lets users query multiple models simultaneously. The unit economics of offering such multi‑model queries at a flat subscription rate remain unclear, but executives believe token allocation to the best‑fit model will keep margins healthy.

Upcoming Products and Events

  • Comet browser for iOS is slated for release next month.
  • A developers’ conference, Ask, is planned for March 11 in San Francisco to promote third‑party use of Perplexity’s API.

User Feedback and Rate Limits

The Perplexity subreddit has reported new rate limits on both free and subscription tiers. Executives dismissed these complaints:

“Any discussions on the free tier being made worse or rate‑limited is completely false,” an executive said.

Despite the criticism, the company’s leadership says they are now tracking revenue metrics more closely than daily query counts, indicating a stronger focus on the bottom line.

0 views
Back to Blog

Related posts

Read more »