Power your products with web-wide research, Q&A capabilities
Perplexity's API now covers the full agent stack: multi-provider model access, real-time search on 200B+ URLs, and SOTA embeddings. One key, one bill, direct provider pricing. Built for developers who are tired of stitching together five vendors to ship one thing.
Perplexity MCP Server connects AI assistants (Claude, Cursor, VS Code) to four tools: perplexity_search, perplexity_ask, perplexity_research, and perplexity_reason. One-click installation enables real-time web search and advanced reasoning directly within MCP-compatible workflows without custom integration code.
Agent API provides unified access to OpenAI, Anthropic, Google, and xAI models with integrated web search, fetch_url, people_search, and finance_search tools. Users can orchestrate multi-provider models with fallback chains and presets for common configurations, enabling autonomous agents with transparent per-token pricing.
Search API returns raw, ranked web search results with advanced filtering, domain controls, and multi-query support at $5 per 1K requests. No token costs—ideal for building custom search experiences, data collection, or feeding results into proprietary models.
Sonar API provides Perplexity's optimized models (Sonar, Sonar Pro, Sonar Deep Research) with built-in citations, streaming support, and conversation memory. Pro Search enables multi-step reasoning with automated tool usage for complex queries, priced per token plus request fees based on search context depth.
Embeddings API generates standard and contextualized embeddings in 1024 or 2560 dimensions for semantic similarity, retrieval-augmented generation, and text classification. Priced at $0.004-$0.05 per 1M tokens, enabling efficient vector-based search and recommendation systems.