Exa Research

AI-powered web search and content retrieval using Exa API.

Scripts Location

./scripts/
├── exa-search    # Search the web
├── exa-contents  # Get page contents
└── exa-similar   # Find similar links

Core Workflows

1. Basic Search

./scripts/exa-search "your query"

Search Types:

• auto (default) - Intelligently combines methods
• fast - Quick results, lowest latency
• deep - Comprehensive with query expansion
• neural - AI semantic search

Examples:

# Quick search
exa-search "React 19 new features"

# Deep research
exa-search "state of LLM agents 2025" --type deep --text

# News only
exa-search "OpenAI GPT-5" --category news --num 5

# Domain filtered
exa-search "machine learning tutorials" --include-domains arxiv.org,paperswithcode.com

2. Get Page Contents

./scripts/exa-contents "https://example.com" [options]

Examples:

# Get text content
exa-contents "https://blog.example.com/post" --text

# Get summary
exa-contents "https://paper.example.com" --summary

# Multiple URLs with LLM context
exa-contents "https://url1.com" "https://url2.com" --context

# Crawl subpages (for docs)
exa-contents "https://docs.example.com" --subpages 5

3. Find Similar Links

./scripts/exa-similar "https://example.com" [options]

Examples:

# Find competitors
exa-similar "https://stripe.com" --num 10

# Similar articles with content
exa-similar "https://blog.example.com/ai-post" --text

# Filter by domain
exa-similar "https://openai.com" --include-domains techcrunch.com,wired.com

Common Options Reference

Research Workflow Patterns

Pattern 1: Topic Research

# 1. Broad search with deep mode
exa-search "quantum computing applications 2025" --type deep --text | jq '.results[:5]'

# 2. Get detailed content for top results
exa-contents "https://result-url.com" --summary --context

Pattern 2: Competitive Analysis

# 1. Find similar companies
exa-similar "https://target-company.com" --num 10

# 2. Deep dive on specific competitor
exa-search "CompetitorName product features" --category company --text

Pattern 3: News Monitoring

# Recent news only
exa-search "AI regulation" --category news --start-date "2025-01-01" --num 20

Pattern 4: Technical Documentation

# Search for docs
exa-search "Next.js App Router caching" --include-domains nextjs.org,vercel.com --text

# Or crawl docs site
exa-contents "https://nextjs.org/docs/app/caching" --subpages 3 --context

Response Processing Tips

# Extract URLs only
exa-search "query" | jq -r '.results[].url'

# Get titles and URLs
exa-search "query" | jq '.results[] | {title, url}'

# Extract context string for LLM
exa-search "query" --context | jq -r '.context'

# Check cost
exa-search "query" | jq '.costDollars'

Error Handling

Categories Reference

• company - Business websites
• research paper - Academic papers
• news - News articles
• pdf - PDF documents
• github - GitHub repositories
• tweet - Twitter/X posts
• personal site - Personal blogs/sites
• financial report - SEC filings, reports
• people - People profiles