Exa AI
exa.ai
Neural search engine designed to fetch clean, structured data for AI agents. Uses embeddings-based search rather than keyword matching, making it far more effective at finding semantically relevant content — paired with a contents API for full-text extraction.
More in Agent Tools, Perception & Capabilities
AI-optimized search API returning cited, summarized answers alongside traditional results. Purpose-built for grounding LLM responses with current web content. Supports RAG snippets and code search modes.
Cloud infrastructure for giving LLMs the ability to use tools with a universal tool store. Provides pre-built tools for web search, code execution, email, and more as MCP or function-call compatible APIs. Handles auth, caching, and rate limiting.
Open-source browser infrastructure for AI agents, providing scalable browser sessions with built-in proxies and captcha solving. Acts as a managed Puppeteer/Playwright cluster with session persistence. Self-hostable on any cloud.
Screenshot API for agents that need visual context of web pages. Renders any URL to PNG or PDF with custom viewport sizes, full-page capture, and element blocking. Useful for agents that reason about page layouts.
Sonar API from Perplexity that provides LLM completions grounded in real-time web search results. Returns cited responses with source URLs. Ideal for agents that need current, factual answers without building a separate search stack.
Visual workflow automation platform with 1800+ app integrations. Connects AI agent outputs to downstream business tools through a drag-and-drop scenario builder. Supports complex branching logic and data transformation.