Perplexity

Perplexity

Overview

Perplexity is a San Francisco-based AI company founded in 2022 by Aravind Srinivas (CEO), Denis Yarats, Johnny Ho, and Andy Konwinski, focused on AI-native search. Rather than competing as a frontier base-model lab, Perplexity built a product layer — an "answer engine" that combines real-time web retrieval with LLM synthesis — and has used that product surface to grow into a $20B+ valuation by early 2026. Perplexity's Sonar models, while based on Meta's Llama foundation, are tuned and operated by Perplexity for the specific synthesis-with-citations workflow that defines the product.

Perplexity reached a $21.21B valuation in early 2026 after closing a Series E-6 round, with annualized revenue growing from ~$80M in late 2024 to ~$200M+ by February 2026. The company committed $750M to Microsoft Azure infrastructure in January 2026, signaling aggressive scale-up. Perplexity is also at the center of ongoing copyright legal battles with publishers over how its answer engine cites and synthesizes news content — a dispute that has become one of the defining legal questions for AI-native search.

Key Details

  • Founded: 2022
  • Co-founders / CEO: Aravind Srinivas (CEO), Denis Yarats, Johnny Ho, Andy Konwinski
  • Headquarters: San Francisco, CA
  • Funding: $1.22B+ cumulative (Series E-6 in early 2026)
  • Valuation: ~$21.21B (early 2026)
  • 2026 ARR: ~$200M (vs. ~$80M late 2024)
  • Distribution Model: Perplexity consumer app + Sonar API + enterprise search
  • Website: https://www.perplexity.ai

Current Models

  • Sonar / Sonar Pro — Perplexity's primary engine; based on Meta's Llama 3.3 with proprietary tuning and retrieval orchestration
  • Sonar Reasoning — Reasoning-tuned variant
  • Sonar Online Models — Continuously web-grounded models for real-time search
  • API Pricing: $1 per million tokens (in/out) for base Sonar; $3 input / $15 output for Sonar Pro

Key People

  • Aravind Srinivas — CEO and co-founder; OpenAI alumnus; one of the most public faces of the AI-native search thesis

Recent Developments

  • $21B+ Valuation (early 2026): Series E-6 closed at $21.21B post-money — Perplexity is now in the same valuation tier as Mistral and ahead of most application-layer AI companies.
  • $750M Azure Commitment (January 2026): Microsoft Azure infrastructure investment signaling aggressive scaling of search infrastructure.
  • Revenue Growth: ARR roughly tripled from $80M in late 2024 to ~$200M by February 2026, validating the consumer-search subscription thesis at scale.
  • Pricing Structure: Pro at $20/month for individuals; Enterprise at $40/user/month; Sonar API at $1 (base) to $3/$15 (Pro) per million tokens.
  • Copyright Litigation: Ongoing publisher lawsuits remain a defining business and legal risk — Perplexity's answer-with-citations approach has been contested as substituting for source publishers' own audiences.

Why They Matter

Perplexity is the canonical case for "the application layer is where the value is" in AI. Without training a single frontier model — Sonar runs on top of Meta's Llama foundation — Perplexity built a $21B company by being the first to ship the AI-native search product that actually feels different from Google. The company's growth trajectory ($80M → $200M ARR in roughly 18 months), $750M Azure infrastructure commitment, and continued investor support validate that AI-native search is a real and durable product category, not a feature that Google or Bing will simply absorb. Perplexity also matters as a legal precedent vehicle: the publisher litigation around how Perplexity cites news content will likely set precedent for how AI-native search products of all kinds handle copyrighted source material — making Perplexity's outcomes consequential well beyond the company itself.

Last Updated

May 7, 2026

→ Back to Companies