Perplexity
An AI answer engine with real-time web sources — ask anything, get cited answers.
Perplexity combines a large language model with live web search to answer questions with up-to-date, cited sources. It is ideal for research, fact-checking and staying current, and it can also follow up conversationally or summarize a whole topic with its Pro Search mode.
Key Features
- Live web search with citations
- Pro Search (deep research)
- Follow-up conversations
- Collections & libraries
- No hallucinated URLs — sources shown
Ratings & Reviews
Loading reviews…
Prompts & Skills for Perplexity
Skills
Chain-of-Thought — let the model reason step by step
Ask the AI to think out loud and solve problems in steps to dramatically improve accuracy on reasoning tasks.
RAG Prompting — retrieval-augmented generation for grounded answers
Combine retrieved context with targeted prompts so the AI answers from your documents instead of hallucinating.
Chain-of-Density — progressively denser, entity-rich summarization
Generate summaries that iteratively incorporate more entities and details while staying concise, producing information-dense summaries without fluff.
Related Tools
Paper Search MCP
Search, analyze, and download academic papers from 25+ sources. MCP server + CLI + Agent Skill.
Paper Search MCP is a comprehensive academic paper search, analysis, and download tool available as an MCP server, CLI, and Agent Skill. Search across 25+ academic sources including arXiv, Semantic Scholar, Crossref, OpenAlex, PubMed, IEEE, ACM, Springer, Elsevier, and more. Features include semantic search, citation analysis, paper metadata extraction, PDF download, reference parsing, and paper summarization. The MCP server integrates with Claude Desktop, Cursor, Cline, and any MCP-compatible client. The CLI provides fast command-line access. The Agent Skill works with Claude Code, Codex, and other AI coding agents.
SwarmVault
Build knowledge graphs from any content. PDF, Word, video, YouTube → graph database. Semantic search over your knowledge.
SwarmVault is a knowledge graph construction tool that transforms unstructured content into structured, queryable knowledge graphs. Feed it PDFs, Word documents, videos, YouTube URLs, or text, and it extracts entities, relationships, and events to build a graph database. Features include multi-document ingestion, entity resolution, relationship extraction, semantic search, graph visualization, and natural language querying. Perfect for researchers, analysts, teams managing large document collections, and anyone who wants to discover hidden connections in their knowledge. Supports multiple graph backends and export formats.
DeepTutor
Agent-native lifelong personalized tutoring workspace by HKUDS — chat tutoring, problem solving, quiz generation, research and mastery practice in one runtime.
DeepTutor is an agent-native learning workspace that connects tutoring, problem solving, quiz generation, research, visualization, and mastery practice in one extensible system. Built by HKUDS (Hong Kong University Data Science). Features a unified runtime for chat, ask-questions, quiz, research, visualize, solve, mastery path and immersive reading. Supports multi-engine RAG (LlamaIndex, GraphRAG, LightRAG, PageIndex), subagent consultation (Claude Code, Codex, Grok CLI), persistent IM partners across Feishu/Telegram/Slack/Discord, immersive YouTube/Bilibili video learning with transcript-grounded tutoring, and the EduHub community skill registry.