Google Images Scraper — Original URLs, Source & Size

Scrape Google Images by keyword and get back the original full-resolution image URL, the source page, a thumbnail, the title, pixel dimensions, and a product flag — as clean structured JSON. Multi-query, 40+ countries, with size / type / color / usage-rights filters. HTTP-only via a managed SerpApi backend, no login, no cookies, no browser. MCP-ready for AI agents.

What it does

Give it one or more search keywords. For each query it returns up to 500 images, each as a flat record with the direct link to the original file plus the page it was found on — ready to download, dedupe, or feed into a dataset.

When to use it

• Build image datasets for machine-learning / computer-vision training.
• Collect product images and competitor visuals for ecommerce.
• Source design inspiration, mood boards, and reference imagery at scale.
• Pull press / brand imagery for marketing and SEO research.
• Power an AI agent that needs to find images by description (MCP tool).

Do not use this to scrape a specific website's gallery (use a site crawler) or to bypass image licensing — see the legal note below.

Output

One record per image:

Field	Description
query	The search query the image was returned for
position	1-based rank within the query
title	Image title / alt text
imageUrl	Direct URL to the original full-resolution image
thumbnailUrl	Small Google-hosted preview thumbnail
sourceName	Website hosting the image
sourcePageUrl	Page the image appears on
originalWidth / originalHeight	Pixel dimensions, when known
isProduct	true if Google tagged it as a shoppable product
scrapedAt	ISO 8601 timestamp

{
  "query": "golden retriever puppy",
  "position": 1,
  "title": "Golden Retriever Puppy sitting in grass",
  "imageUrl": "https://example.com/images/golden-retriever-puppy.jpg",
  "thumbnailUrl": "https://encrypted-tbn0.gstatic.com/images?q=tbn:...",
  "sourceName": "example.com",
  "sourcePageUrl": "https://example.com/puppies/golden-retriever",
  "originalWidth": 1920,
  "originalHeight": 1280,
  "isProduct": false,
  "scrapedAt": "2026-06-30T12:00:00.000Z"
}

Pricing

Pay-per-event:

• Actor start: $0.00005 per run
• Image scraped: $0.002 per image returned

A 100-image query costs about $0.20. The cost cap is printed in the run log before any charge fires, and the actor never charges beyond Max results per query × number of queries.

Input

Field	Required	Default	Notes
queries	✅	—	List of keyword searches, e.g. ["golden retriever puppy"]
maxResultsPerQuery		100	1–500, paginates automatically
country		us	2-letter ISO code (gl)
language		en	2-letter code (hl)
safeSearch		off	off or active
imageSize		any	large / medium / icon
imageType		any	photo / clipart / lineart / gif / face
imageColor		any	color / grayscale / transparent
usageRights		any	creativeCommons / commercial

Example input

{
  "queries": ["vintage neon sign", "art deco poster"],
  "maxResultsPerQuery": 100,
  "country": "us",
  "imageType": "photo",
  "imageSize": "large"
}

Use it from the API

curl -X POST "https://api.apify.com/v2/acts/khadinakbar~google-images-scraper/runs?token=YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"queries":["golden retriever puppy"],"maxResultsPerQuery":50}'

from apify_client import ApifyClient

client = ApifyClient("YOUR_TOKEN")
run = client.actor("khadinakbar/google-images-scraper").call(
    run_input={"queries": ["golden retriever puppy"], "maxResultsPerQuery": 50}
)
for item in client.dataset(run["defaultDatasetId"]).iterate_items():
    print(item["imageUrl"])

MCP / AI agents

Exposed as apify--google-images-scraper in the Apify MCP server. Pass a keyword in, get structured image records out — ideal for agents that need to find or collect images by description.

FAQ

Do I need a SerpApi key? No. The backend is fully managed — there is nothing to configure.

How many images per query? Up to 500. Google serves roughly 100 per page and the actor paginates automatically.

Are the image URLs the originals or thumbnails? imageUrl is the original full-resolution file. thumbnailUrl is the small Google preview.

Why is an image's imageUrl sometimes null? A small fraction of Google results omit the original; the thumbnail and source page are still returned.

Legal

This actor returns links to publicly visible images and their source pages. It does not grant any license to those images. The usageRights filter reflects Google's labeling and is not legal clearance — you are responsible for verifying the license on each image's source page before downloading, reusing, or redistributing it. Use in compliance with Google's Terms of Service and applicable copyright law.