Perplexity
Identity card
The company
- Name
- Perplexity ↗
- Founded
- 2022
- Head office
- San Francisco, California, United States
- Chief executive
- Aravind Srinivas
The product
- Flagship model
- Sonar
- Subscription
- €22 per month · Pro, on the App Store
- API
- 0.25 $ / 2.50 $ input / output, per million tokens
The data
- Law
- United States
Usage
- Web traffic, Sept. 2026
- 0.9 %
Key points
Perplexity is an answer engine: each answer links to the pages read, which makes it a popular tool for documentary research. The reliability of these answers remains uneven, and several major media outlets are suing the company for reusing their articles.
- Search with sources
- Monitoring
- Long-form writing
- Believing the news without checking
Source: Similarweb, share of visits to generative AI sites, worldwide; records published by Similarweb and repeated by the press.
Yes for looking up information and seeing where it comes from. Look elsewhere for writing, analysing a long document or coding (Claude, ChatGPT).
Strengths
- Sources with every answerEach question launches a web search, and the answer cites the pages it comes from: you can check in one click.Source: Wikipedia, 7 November 2026 ↗
- One of the cheapest APIsIts Sonar model costs $0.25 per million input tokens, the lowest price among the flagship models in the ranking.Source: Perplexity, API documentation, 5 November 2026 ↗
- A browser that actsComet, launched in July 2025 and free since October 2025, browses and completes tasks for the user.Source: Wikipedia, 7 November 2026 ↗
Weak points
- Errors on the newsNewsGuard found 47% false claims on news topics in August 2025, the 2nd worst result of the 10 tools tested.Source: NewsGuard, August 2025 ↗
- Sued by media outletsThe New York Times and the Chicago Tribune sued it in December 2025 for reusing their articles; Amazon has been suing it since November 2025 over its shopping agent.Source: Engadget, 5 December 2025 ↗ ; Bloomberg, 4 November 2025 ↗
- Little used on the web0.9% of the web traffic of AI sites in September 2026, a share stable or declining for several months according to Similarweb data.Source: OfficeChai, citing Similarweb, 12 October 2026 ↗
The company Perplexity
The Perplexity entry →Company history
Perplexity was founded in San Francisco in August 2022 by Aravind Srinivas, Denis Yarats, Johnny Ho and Andy Konwinski. Their idea: replace a search engine's page of links with a written, sourced answer.
The story continues
The engine opens in December 2022. Perplexity relies on other companies' models and on its own, launches the Comet browser in 2025 and in August 2025 offers to buy Chrome for 34.5 billion dollars, an offer that went nowhere.
Its valuation reaches 20 billion dollars in September 2025, while media lawsuits multiply.
Sources: Wikipedia, 7 November 2026 ↗ ; Seeking Alpha, citing The Information, 10 September 2025 ↗
Leadership
- Chief executive officer and co-founder
- Aravind Srinivas ↗
- Co-founders
- Denis Yarats, Johnny Ho, Andy Konwinski ↗
Key figures
- Valuation, September 2025
- 20 billion dollarsSeeking Alpha, citing The Information, 10 September 2025 ↗
Perplexity plans
- FreeFree0 €0 $
- ProStandard22 €20 $
- MaxPremiumprice published in dollars only200 $no euro price
- Enterprise ProEnterprise, per seatprice published in dollars only40 $no euro price
Perplexity pricing pages: including VAT in France, excluding tax in the United States; up to date as at 9 November 2026.
Content of the 4 plans, prices in France and the United States
Prices in France and the United States. In the United States, prices are shown in dollars and excluding tax; in France, they include 20% VAT. So $20 does not equal €20: the French price is often slightly higher. When only the dollar price appears, the company does not publish a euro price: the amount paid in France depends on the exchange rate and VAT.
VibeMistral Large1.50 $0.50 $- PerplexitySonar2.50 $0.25 $1.7 times the output price of Vibe
DeepSeekV4-Pro3.96 $1.32 $
Meta AIMuse Spark 1.14.25 $1.25 $
QwenQwen3.8-Max6 $2 $
GrokGrok 4.76 $2 $
GeminiGemini 4 Argon10 $2 $
ClaudeOpus 5.520 $4 $
ChatGPTGPT-6 Astra50 $10 $
Copilot: no API of its own. Flagship model of each company, standard rate, up to date as of 9 November 2026.
Prices up to date as of 9 November 2026, read on the companies' pages, or from 2 concordant sources; the date of each record is in the page's price file Methodology.
Perplexity models
- Company
- Perplexity
- Models offered
- "Best" (automatic model choice)by default, Free and paid plans (official page ↗)GPT-6.1 Solpaid plan, Pro, Max; 3rd of 71 for general intelligence (official page ↗)GPT-6 Astrapaid plan, Max; 2nd of 71 for general intelligence (official page ↗)Gemini 4 Argonpaid plan, Max; 7th of 71 for general intelligence (official page ↗)Gemini 3.8 Flashpaid plan, Pro, Max; 22nd of 71 for general intelligence (official page ↗)Claude Sonnet 5.5paid plan, Pro, Max; 8th of 71 for general intelligence (official page ↗)Claude Opus 5.5paid plan, Max; 4th of 71 for general intelligence (official page ↗)Claude Fable 5.1paid plan, Max; 6th of 71 for general intelligence (official page ↗)Grok 4.7paid plan, Pro, Max; 36th of 71 for general intelligence (official page ↗)Kimi K3 (Moonshot AI), GLM 5.3 (Z.ai), Nemotron 3.5 Super VL (NVIDIA)paid plan, Pro, Max (official page ↗)
Links read on 10/11/2026 on the company's pages, press as a fallback. Rank: Epoch AI ECI index (CC BY 4.0). Performance of AI models
Data management
How Perplexity handles conversations
- Training models on conversationsFree, Pro and Max plans: yes by default ("AI data retention" setting). Enterprise Pro and Enterprise Max: never.Source: Perplexity, help centre, 9 November 2026 ↗
- Opting out of trainingTurn off "AI data retention" in the account settings, under Preferences; this applies to subsequent data.Source: Perplexity, help centre, 9 November 2026 ↗
- Applicable lawUS law: Perplexity is a United States company, subject in particular to the Cloud Act, even when the servers are in Europe.Source: Guide d'usage de l'IA pour les agents publics de l'État (DINUM), June 2026 ↗
Score on this criterion: 33 out of 100, joint 4th out of 10. Retention for AI active by default, can be turned off; stated period contradictory; no procedure in Europe. The ethics ranking
Environment
Environment score
42 6th of 10- Simple query 40 %50
- Advanced queries 20 %58
- Decarbonised electricity 20 %33
- Environmental approach 20 %17
Black line: average of the 10. The environmental ranking
Our view
Each query to Perplexity launches a web search and the reading of several pages, with a model chosen without being displayed, often Sonar, of medium size. Heavy uses remain limited: no image or video highlighted, but Deep Research and the Comet agent browser. Perplexity rents its servers from AWS and Azure without publishing the countries or the electricity, and publishes no figure and no report.
Per-query figure published by Perplexity
Nothing publishedPerplexity publishes no per-query figure, either for energy or for emissions. 3 of 10 companies publish one: Google, Mistral AI and OpenAI.
Company figures cannot be compared with each other: scopes and methods differ. The "Simple query" criterion of the environmental ranking combines them with external estimates, for lack of a common basis: the methodology sets out the limits.
- Gemini0.24 Wh
Copilotnothing published- Vibe1.14 gCO2e
- Claudenothing published
- ChatGPT0.34 Wh
- Qwennothing published
- Meta AInothing published
- Perplexitynothing published
- DeepSeeknothing published
- Groknothing published
On the Éclairages blog: What is the electricity consumption of AI? ↗ ; Which AIs pollute the most? ↗
Scores and criteria
The ranking →The Celsius score is the average of 3 scores: performance, ethics and environment. Each combines several criteria. Each row places Perplexity among the 10; clicking a criterion opens its definition and the justification for Perplexity's score.
5 public rankings 100 %289th of 10
Definition. Synthesis of 5 public performance rankings, where models are ranked by tests or by votes from internet users: each ranking gives 100 to the best of the 10 and 0 to the last.
Justification. No model of its own: estimate, average of its 2 neighbours in the ranking, Grok and Vibe.
Content produced 30 %58joint 4th out of 10
Definition. Content the assistant produces on request (sexual, misleading images, minors) and results of independent tests.
Justification. Pornography prohibited, no known sexual incident; no deepfake rules and no age verification; 47% false claims (NewsGuard, 8/2025).
Use of data 30 %33joint 4th out of 10
Definition. Use of conversations for training, retention period, storage location, sanctions by the authorities.
Justification. Retention for AI active by default, can be turned off; stated period contradictory; no procedure in Europe.
Safety 20 %257th out of 10
Definition. Independent safety evaluations (Future of Life Institute, SaferAI, Phare) and resistance to attacks.
Justification. No safety framework, no Seoul commitments, no European code; absent from independent evaluations.
Governance 20 %33joint 5th out of 10
Definition. Control of the company and deliberate choices: military uses, shareholding, commitments kept or abandoned.
Justification. Serves other companies' models; sued by at least 6 media outlets and platforms, accused by Cloudflare of getting around blocks (8/2025).
Simple query 40 %50joint 5th out of 10
Definition. Consumption of a typical text query, on the default model served: figure published by the company, third-party estimate, failing that the size of the model.
Justification. Each question triggers a web search and the reading of pages; model chosen without being shown (often Sonar, medium size).
Advanced queries 20 %58joint 3rd out of 10
Definition. Weight of the resource-hungry uses the product promotes: video and image generation, agents, reasoning enabled by default.
Justification. Deep Research and the Comet agent browser; no image or video highlighted.
Decarbonised electricity 20 %33joint 4th out of 10
Definition. Share of low-carbon electricity in the data centres and efforts to increase it: local grid, on-site gas, renewable or nuclear electricity contracts.
Justification. Rented clouds (AWS, Azure), countries and electricity not published.
Environmental approach 20 %179th of 10
Definition. What the company publishes and commits to: energy and emissions figures, environmental report, targets, model optimisation.
Justification. No figure, no report, no published optimisation.
Compare
or Perplexity or ChatGPTPerplexity searches and cites; ChatGPT does everything else, and searches too.Celsius score: 36 against 62 or Perplexity or GeminiPerplexity cites its sources with every answer; Gemini relies on Google search and is better rated on environment.Celsius score: 36 against 71 or Perplexity or ClaudePerplexity for finding, Claude for reading, writing and analysing.Celsius score: 36 against 74Other comparisons
Perplexity or Copilot−22Perplexity or Meta AI−12Perplexity or Vibe−4Perplexity or Qwen−3Perplexity or DeepSeek+1Perplexity or Grok+10Celsius score gap.
Ranked 8th of 10 in the assistant ranking, scores as at 11 November 2026. Facts and figures in the entry collected from 7 to 9/11/2026, each with its source; foundation and headquarters: Perplexity AI, Form D (SEC, 5/4/2023), 11 November 2026 ↗.
Continue in the ranking
The full ranking →CompanyPerplexityNo model of its own measured; Perplexity serves those of OpenAI, Google, Anthropic and SpaceX Super Intelligence.The entry →ModelOpenAIGPT-6 Astra2ndout of 71 on general intelligenceThe best ranked for general intelligence among the models Perplexity offers.The entry →Comparison2 AI assistantsPerplexity or ChatGPT36 against 62Celsius score out of 100Perplexity searches and cites; ChatGPT does everything else, and searches too.The comparison →The hub newsletter
This month's ranking, news on the impact of AI and one figure from the AI Impact Database. One newsletter a month at most, free of charge. Past issues
