Domain Presets: Optimized Sources for Every Industry

The Generic Search Problem

Generic web search returns mixed results:

• News articles when you want prices
• Blog posts when you want academic papers
• Forums when you want official docs
• Low-quality sites mixed with high-quality sources

Result: Noise, irrelevant results, wasted time.

The Domain Preset Solution

Deep Research API provides curated source lists for specific domains:

Available Presets

Crypto

const cryptoData = await deepResearch({
  query: "Bitcoin price analysis",
  preset: "crypto"
})

// Searches:
// - CoinMarketCap, CoinGecko
// - CoinDesk, CryptoSlate
// - Exchange APIs (Binance, Coinbase)
// - Blockchain explorers

Stocks

const stockData = await deepResearch({
  query: "AAPL financials",
  preset: "stocks"
})

// Searches:
// - Yahoo Finance, Google Finance
// - SEC filings, EDGAR
// - Bloomberg, Reuters
// - Company investor relations

Tech

const techInfo = await deepResearch({
  query: "React 19 new features",
  preset: "tech"
})

// Searches:
// - Official docs (React.dev, MDN)
// - GitHub repositories
// - Stack Overflow, Dev.to
// - Tech blogs, Medium articles

Academic

const academicPapers = await deepResearch({
  query: "transformer architecture improvements",
  preset: "academic"
})

// Searches:
// - arXiv, PubMed
// - Google Scholar
// - ResearchGate, Semantic Scholar
// - University repositories

News

const newsData = await deepResearch({
  query: "AI regulation updates",
  preset: "news"
})

// Searches:
// - Reuters, AP, Bloomberg
// - TechCrunch, The Verge
// - Industry publications
// - Press releases

How Presets Work

Source Curation

Each preset has 20-50 curated sources:

const PRESETS = {
  crypto: [
    "coinmarketcap.com",
    "coingecko.com",
    "coindesk.com",
    "binance.com",
    "coinbase.com",
    // ... 45 more
  ],
  stocks: [
    "finance.yahoo.com",
    "google.com/finance",
    "sec.gov",
    "bloomberg.com",
    // ... 40 more
  ],
  tech: [
    "react.dev",
    "developer.mozilla.org",
    "github.com",
    "stackoverflow.com",
    // ... 35 more
  ]
}

Quality Filtering

Sources are vetted for:

• Authority: Domain authority score
• Freshness: Content update frequency
• Relevance: Topic alignment
• Reliability: Historical accuracy

Benefits

Faster Results

Curated sources = fewer low-quality results:

// Without preset: 50 sources, 30% low quality
// Processing time: 45 seconds

// With preset: 25 sources, 95% high quality
// Processing time: 30 seconds

Better Accuracy

Domain-specific sources provide better data:

// Generic search might find:
"According to a Reddit post, Bitcoin is $95,000..."

// Crypto preset finds:
"Bitcoin (BTC) is trading at $95,000 on CoinMarketCap..."

Reduced Noise

Fewer irrelevant results:

// Generic search for "React hooks":
// - Blog posts about hooks (fishing)
// - React component libraries
// - Unrelated tech articles

// Tech preset for "React hooks":
// - React.dev official docs
// - React hooks tutorial sites
// - Stack Overflow answers
// - GitHub examples

Custom Presets

Create your own source lists:

const customPreset = await deepResearch({
  query: "Your query",
  sources: [
    "your-domain-1.com",
    "your-domain-2.com",
    "your-domain-3.com"
  ]
})

Get Started

Get better results with domain presets.

Get Your API Key

Read Preset Documentation