exa-search

Use this when the task begins with discovery rather than known URLs.

Best Uses

• find authoritative sources for a topic
• discover entities, people, companies, products, or docs pages
• build candidate citation sets before extraction
• rank likely-useful pages for downstream summarization

Required Env

• EXA_API_KEY

Output Priorities

Return a compact evidence set:

• title
• url
• snippet
• reason selected
• confidence

Do not fetch full page content unless the task explicitly needs it or the runtime has a lightweight Exa contents endpoint wired.

Query Patterns

import os
import httpx

EXA_API_KEY = os.environ["EXA_API_KEY"]

async def exa_search(query: str, num_results: int = 5) -> dict:
    async with httpx.AsyncClient(timeout=30) as client:
        resp = await client.post(
            "https://api.exa.ai/search",
            headers={
                "x-api-key": EXA_API_KEY,
                "Content-Type": "application/json",
            },
            json={
                "query": query,
                "numResults": num_results,
                "useAutoprompt": True,
                "type": "auto",
            },
        )
        resp.raise_for_status()
        return resp.json()

Selection Heuristics

Prefer sources that are:

• primary or official when the task is factual
• recent when the task is temporal
• direct documentation when the task is technical
• high-signal niche sources over generic SEO pages

Avoid

• deep extraction work when URLs are already known
• hostile surfaces that need rendering or proxying
• SERP capture tasks

Hand-off Rules

• Pass chosen URLs to firecrawl-extract for content capture.
• Escalate to brightdata-collection only if discovery must happen on a blocked or geo-specific surface.