Exa is a search engine built for AIs. We enable developers to filter the internet's content with complex queries in order to bring the world's knowledge into their AI apps.
The highest quality web search API for AI, returning relevant, fresh results at every latency.
What customers say
“Exa was better than the native model search experience for us.”
— Prakash Balasubramanian, HubSpotExa customer
“We're entering a new era where software actually does the work. For that to be real, the agents need live intelligence behind them. Exa is where that comes from: current company and people data sourced from the web in real time.”
CustomerExa customer
“Exa is necessary for agents on HubSpot to supplement internal data with high quality, real-time people and company search. We found that other alternatives were slower, more expensive, and did not have comprehensive enrichment coverage.”
CustomerExa customer
Case study
Obvious uses Exa's Websets API to build market maps and identify contacts. Exa reports that one enriched call replaces 15–20+ manual research iterations and substantially reduces market-intelligence costs.
CodeRabbit uses Exa Search and Deep Search to verify code-review comments against current documentation, packages, and public repositories. Exa reports 70–75% fewer search calls and verification latency of 7.1 seconds…
OpenRouter uses Exa as a web-search engine, passing relevant highlights and citations to models through its server-side search tooling. Exa reports 73 million search queries across OpenRouter's web-connected model…
Exa Launches MCP Server for AI Assistant Integration
Exa has launched the Exa MCP (Model Context Protocol) Server, enabling AI assistants to directly access Exa's search, web-fetching, and agent capabilities. The integration supports numerous development environments and AI tools, including Claude, Cursor, VS Code, and Grok. The MCP server allows for standardized connection and configuration of Exa's tools across various client applications, facilitating features such as web search, advanced filtering, and multi-step agent operations. Users can configure permissions, enable optional advanced tools, and manage API keys for production use through the client-side configuration files.
IntegrationClaudeIntegrationCursorIntegrationVS CodeIntegrationGrokIntegrationWindsurfIntegrationGoogleIntegrationZedIntegrationVercel Inc.
Product launch
Exa Launches Agent for Deep Research and Structured Data Enrichment
Exa has introduced Exa Agent, an asynchronous endpoint designed for deep research, list-building, and data enrichment workflows. The agent handles multi-step reasoning tasks and provides structured JSON outputs based on custom schemas. It includes capabilities for schema-validated results, field-level grounding, and metadata generation. Exa Agent supports integration with premium data partners, such as Similarweb, allowing users to select specific providers for data enrichment tasks. The service features usage-based pricing, with variable effort modes ranging from minimal to extra-high to balance cost and output complexity, and is accessible via the Exa SDK and OpenAI-compatible API.
