build-with-exa

On Activation

1. Confirm EXA_API_KEY is available for any live API call. SDK examples assume the key is set.
2. Default new integrations to POST /search with type: "auto" and contents.highlights: true. Escalate to Agent API only for multi-step research / list-building.
3. Load only the references/*.md file needed for the current endpoint - do not dump the whole tree into context.
4. For agent-native marketing research inside mktg projects, prefer exa-search, company-research, or lead-generation over inventing a custom integration.

Build with Exa

Scope

Included by default:

• Core retrieval APIs: search endpoint, contents endpoint, answer endpoint, context endpoint
• Long-running research workflows: Agent API (/agent)
• Async and recurring workflows: Monitors API, Websets API
• SDK guidance: Python exa-py, TypeScript exa-js

Note on data retention: /search, /answer, and deep research are Zero Data Retention (ZDR). The Agent API (/agent), Websets, and Monitors are not ZDR. If a use case requires ZDR, stay on the ZDR surfaces or contact Exa.

Installation

# Python
pip install exa-py

# TypeScript / JavaScript
npm install exa-js

Authentication

export EXA_API_KEY="your_api_key_here"

Exa accepts either the x-api-key header or Authorization: Bearer <key>.

API Decision Workflow

Before picking an endpoint, decide which workflow shape fits:

• Raw web content for your own LLM or agent: start with /search using type: "auto" and contents: { highlights: true }
• Synthesized structured output: start with /search using the search type that fits your latency and reasoning needs, then add outputSchema and systemPrompt
• Long-running multi-step research, list-building, or enrichment with structured output: use the Agent API (/agent)

Default to the search endpoint. Use the search endpoint (/search) for most new integrations, then move to a more specialized Exa surface only when the task shape clearly calls for it.