Perplexity
AI-powered answer engine delivering conversational search with real-time results and citations.
What is Perplexity?
Perplexity is a conversational AI-powered search engine—or “answer engine”—that blends real-time web retrieval with large language model (LLM) capabilities. Launched in August 2022 and based in San Francisco, it delivers direct, natural-language responses to user queries, accompanied by transparent source citations. The platform supports both free and subscription-based tiers, with Pro and Enterprise plans offering access to advanced models (e.g., GPT‑4, Claude, Llama‑based Sonar, DeepSeek R1), multimodal input, file uploads, APIs, and team-focused features such as SOC 2 compliance and admin dashboards. The tool has expanded beyond search into products like Perplexity Pages, the Comet browser, and the novel “Personal Computer”—a cloud-based AI assistant running on Mac mini hardware.
What you can do with it
Research assistance with verifiable sourcing
Users conduct in‑depth research using real‑time web search, receiving summaries supported by numbered citations.
Academic coursework support
Students gather and synthesize information, structure writing, and confirm sources using Focus modes and file uploads.
Enterprise knowledge base and onboarding
Teams build searchable onboarding spaces with documentation and wikis accessible via AI‑powered answers.
Content creation and summarization
Creators summarize documents, videos, or articles, aiding workflow with multimodal context.
Automated project workflows
Users set up agents to handle repetitive tasks like memo generation, code review, and workflow execution.
Key features
- Real‑time web retrieval with inline source citations
- Conversational search with follow‑up refinement
- Multiple 'Focus' modes (e.g., Web, Academic, News, Reddit, YouTube, Wolfram Alpha)
- Access to a variety of LLMs (e.g., Sonar, GPT‑5.2, Claude Sonnet, Gemini Pro)
- Deep Research and enhanced query types (chain‑of‑thought reasoning, iterative reports)
- File uploads for contextual queries (documents, code, media)
- Workspaces (‘Spaces’) for organizing threads with custom instructions and collaboration
- Agents and workflow automation integrating web, files, and third‑party apps
Screenshots
Inputs / Outputs
Strengths & Limitations
Strengths
Source transparency
Each answer includes citations, enabling users to verify and explore the underlying information.
Real-time web search
Integrates live information retrieval alongside LLM-generated responses, keeping answers current.
Model flexibility
Pro users can choose among multiple LLMs—including GPT‑4, Claude, Llama‑based Sonar, DeepSeek R1—for diverse perspectives and capabilities.
Multimodal inputs and uploads
Supports file uploads (PDFs, images, code, data), enabling integrated analysis of user documents with web data.
Advanced research tools
Features like Deep Research, Pro Search, and Perplexity Pages provide iterative, detailed reporting capabilities.
Expanding product ecosystem
Includes offerings such as Comet browser and the AI‑powered Personal Computer agent with enterprise integrations.
Limitations
Legal and copyright scrutiny
Has faced lawsuits and cease‑and‑desist letters from major publishers over copyright and content use concerns.
Potential hallucinations
Shares typical generative AI risks like factual inaccuracies and content hallucination.
Free tier usage limits
The free plan restricts advanced (Pro) searches per day, pushing frequent users toward paid plans.
Opaque model orchestration
The system’s internal handling of model choices and instructions can feel like a black box.
Pricing & Plans
Model: Freemium
Free
Unlimited basic searches with real‑time citations, Focus modes, voice search, Collections, Spaces; limited Pro/Deep Research access
Education Pro
Discounted Pro access for verified students
Pro
Unlimited Pro searches, Deep Research reports (~20/month), AI model selection (GPT‑5.2, Claude Sonnet, Gemini Pro), file uploads, image/video generation, API credit
Max
Highest‑usage tier with unlimited Pro searches, Labs features, Comet browser access
Free tier offers unlimited quick searches and limited Pro searches per day. Pro starts at $20/month (or $200/year) with upgraded quotas, file uploads, image generation, and API access. Enterprise plans range up to $325/month per seat for high-volume and enterprise-grade features. Max tier at $200/month for highest usage.
Who it's for
Ideal for
Researchers, professionals, and power users needing accurate, cited information with file integration and advanced model control.
Not ideal for
Casual searchers satisfied with traditional search engines and unwilling to pay for deeper capabilities.
What users say
- Trust via citations
- Impressive for research
- Legal controversy concerns
- Valuable Pro features
Prompts & Results
›Explain quantum entanglement with sources.
Perplexity returns a concise explanation of quantum entanglement along with cited references from educational and scientific websites.
›Upload this PDF and summarize its key findings.
Tool analyzes the uploaded PDF and produces a structured summary tied to specific document sections.
›Compare GPT‑4 versus Claude on sentiment analysis.
Generates a side‑by‑side comparison using different LLMs, noting performance differences and providing citations.
›Perform deep research on climate change policy.
Produces a detailed multi‑source report with structured arguments, context, and cited datasets.
FAQ
What models does Perplexity use?+
Perplexity uses its own Sonar (Llama‑based), DeepSeek R1, and third‑party models like GPT‑4 and Claude, selectable in Pro/Enterprise tiers.
Can I upload my own documents?+
Yes—Pro and Enterprise users can upload PDFs, Word files, images and more to search alongside web content.
Does Perplexity cite its sources?+
Yes—search results include clickable citations for transparency and verification.
What is the difference between free and Pro?+
Free offers unlimited quick searches with limited Pro searches. Pro provides higher limits, model choice, file upload, image generation, and API access.
Ratings & Reviews
No reviews yet — be the first to rate this tool.