Startpage Image Scraper

Search Startpage Images and collect source-backed image records with direct image and thumbnail URLs, source pages, dimensions, format, MIME type, file size, orientation, megapixels, and complete query provenance.

Features

• Single-query and fair multi-query modes
• Native Startpage pagination with overall maxItems
• Width, height, orientation, format, and source-domain filters
• Relevance, largest, and smallest ordering
• Stable IDs, direct URL deduplication, and matched-query tracking
• Startpage embedded structured-data extraction with DOM fallback
• Optional Apify Residential or custom proxies, persistent sessions, consistent browser fingerprints, and bounded retries
• Clean no-results behavior and fail-closed challenge detection

Startpage Images uses a browser page over HTTPS. GOOGLE_SERP is therefore rejected; that proxy group supports specific Google HTTP requests, not Startpage. The Actor does not solve or bypass CAPTCHAs.

Example input

{
  "queries": ["nature wallpaper", "city skyline"],
  "maxItems": 50,
  "maxPages": 2,
  "minWidth": 1920,
  "minHeight": 1080,
  "orientation": "landscape",
  "formats": ["jpeg", "png"],
  "sortBy": "largest",
  "safeSearch": "moderate",
  "proxyConfiguration": { "useApifyProxy": false }
}

Output

Each record contains stable identity and ranking, title, direct image URL, direct thumbnail URL when available, source-page provenance, numeric dimensions, derived aspect ratio/megapixels/orientation, source format and file size when available, extraction method, query/matched-query context, Safe Search, locale, and scrape time. Missing optional values are omitted rather than emitted as nulls.

Startpage-proxy URLs containing ephemeral signatures are not stored. The Actor prefers the source-provided rawImageUrl and decodes proxied thumbnails to their direct image host.

Local development

npm ci
npm test
node validate-datasets.js
apify run --purge --input-file INPUT.json
node scripts/validate-output.mjs storage/datasets/default

Use at least 1 GB of memory; 2 GB is recommended for reliable cloud Firefox startup. Effective browser concurrency is automatically capped to the run memory (roughly one browser per 900 MB) to prevent browser crashes. Search inventory and optional metadata vary by query, locale, network route, and time.