Google Images Scraper — Original URLs, Source & Size

Google Images Scraper by khadinakbar accepts keyword searches and returns one record per image found in Google Images results. Each record can include the original full-resolution image URL, thumbnail URL, source page URL, source name, title, pixel dimensions when known, and a product flag, plus the query and result position that produced it. This Apify Actor is usable through Apify MCP and works with a managed SerpApi backend over HTTP.

Best fit and connected workflows

This Actor fits workflows that start with a keyword and need image-level records for the next step. It is a strong match for:

• image collection for research, taxonomy building, and dataset assembly
• ecommerce visual discovery where the product flag and source page support follow-up review
• design and content research where title, thumbnail, dimensions, and original URL are useful together
• agent workflows that need an Apify tool for image lookup by description through MCP
• Google-centered workflows that continue into Google Maps Reviews Scraper when a returned Google record leads to a business or place and the next step needs reviews
• Use adjacent Google research with Google Ads Scraper - Live Search Sponsored Ads or Google Ads Transparency Scraper - Creatives & Run Dates when the next task is ad-focused rather than image-focused

Practical scenario

Maya is preparing a mood board for a product launch. She starts with the keyword search vintage neon sign and sets imageSize to large so the results favor higher-resolution images. The run returns records with imageUrl, thumbnailUrl, sourceName, sourcePageUrl, originalWidth, originalHeight, and isProduct. Maya uses sourcePageUrl to open the hosting page, checks imageUrl for the direct file, and compares dimensions before choosing which references to include in the board.

Input

Use keyword queries only. Each query is processed independently and can return up to the selected maximum number of images.

Field	Type	Required	Description
queries	array	Yes	One or more keyword searches run on Google Images. Google search operators such as quotes and site: also work.
maxResultsPerQuery	integer	No	Maximum number of images to return per query. Defaults to 100 and can go up to 500.
country	string	No	Two-letter ISO country code used for localized results.
language	string	No	Two-letter UI and result language code.
safeSearch	string	No	SafeSearch setting, either off or active.
imageSize	string	No	Image size filter: any, large, medium, or icon.
imageType	string	No	Image type filter: any, photo, clipart, lineart, gif, or face.
imageColor	string	No	Color filter: any, color, grayscale, or transparent.
usageRights	string	No	Google-reported usage rights filter: any, creativeCommons, or commercial.

Example input

{
  "queries": ["golden retriever puppy", "vintage neon sign"],
  "maxResultsPerQuery": 25,
  "country": "us",
  "language": "en",
  "safeSearch": "active",
  "imageSize": "large",
  "imageType": "photo",
  "imageColor": "color",
  "usageRights": "creativeCommons"
}

Output

The Actor stores a dataset of image records. One record represents one image returned for one query.

Field	Type	Description
query	string	Search query that returned the image.
position	integer	1-based rank within the query results.
title	string or null	Image title or alt text reported by Google.
imageUrl	string or null	Direct URL to the original full-resolution image file.
thumbnailUrl	string or null	Google-hosted preview thumbnail.
sourceName	string or null	Website hosting the image.
sourcePageUrl	string or null	Web page where the image appears.
originalWidth	integer or null	Original image width in pixels when known.
originalHeight	integer or null	Original image height in pixels when known.
isProduct	boolean	Whether Google tagged the image as a shoppable product.
scrapedAt	string	ISO 8601 timestamp when the image was scraped.

Illustrative output record

{
  "query": "vintage neon sign",
  "position": 1,
  "title": "Vintage Neon Sign",
  "imageUrl": "https://example.com/images/vintage-neon-sign.jpg",
  "thumbnailUrl": "https://encrypted-tbn0.gstatic.com/images?q=tbn:example",
  "sourceName": "example.com",
  "sourcePageUrl": "https://example.com/gallery/vintage-neon-sign",
  "originalWidth": 1920,
  "originalHeight": 1280,
  "isProduct": false,
  "scrapedAt": "2026-06-30T12:00:00.000Z"
}

How it works

This Actor uses a managed SerpApi Google Images backend over HTTP. It accepts keyword queries, applies the selected filters, and paginates results up to 500 images per query. The live contract shows a primary charged event named Image scraped, and the dataset view presents the records as a table with query, rank, title, URLs, source, dimensions, and product flag.

Pricing

This Actor uses Apify pay per event pricing plus Apify platform usage. Charges are event-based: one run start event and one Image scraped event for each returned image. For the current live amounts and the platform usage details, check the Pricing tab on this Actor page.

For example, a run that returns one hundred images includes one actor start event and one hundred image events. If you increase the number of queries or the per-query maximum, the number of returned records can grow accordingly.

Use with AI agents (MCP)

Tool description: an Apify Actor tool for keyword-based Google Images lookup that returns structured image records with original URL, source page, thumbnail, title, dimensions, and product flag.

Exact Actor identity: khadinakbar/google-images-scraper

This Actor is usable through Apify MCP, so an agent can send a keyword and receive dataset-backed results with provenance preserved in the returned fields.

Find Google Images for "vintage neon sign" in the US, return the top 20 results, and summarize the matches by source name and image dimensions. Keep the original image URL and source page URL in the records so I can review provenance.

Output interpretation:

• imageUrl points to the original file when Google provides it.
• thumbnailUrl is the Google preview image.
• sourcePageUrl points to the page hosting the image, which helps with provenance checks.
• originalWidth and originalHeight help compare resolution across results.
• isProduct identifies records Google marked as shoppable products.

Scope, pagination, and cost guidance:

• Use keyword queries, including multiple queries in one run.
• Increase maxResultsPerQuery when you want more results from a single query.
• The Actor paginates automatically until the selected per-query maximum is reached, up to 500.
• Cost scales with returned image records through the Image scraped event.

Apify API example

import { ApifyClient } from "apify-client";

const client = new ApifyClient({
  token: process.env.APIFY_TOKEN,
});

const run = await client.actor("khadinakbar/google-images-scraper").call({
  queries: ["golden retriever puppy"],
  maxResultsPerQuery: 10,
  country: "us",
  language: "en",
  safeSearch: "active",
});

const dataset = await client.dataset(run.defaultDatasetId).listItems();
for (const item of dataset.items) {
  console.log(item.imageUrl);
  console.log(item.sourcePageUrl);
}

Best results and outcome guidance

Choose keyword phrases that match the image topic you want to explore. If your task is visual sourcing, imageSize: "large" is a practical first filter. If you want photographs rather than graphics, use imageType: "photo". If you are working with localized results, set both country and language to the region and UI language that fit the workflow. For brand-safe internal use, set safeSearch: "active".

Design note

I found that the dataset contract includes both sourceName and sourcePageUrl, which makes each image record easier to trace back to its hosting site and page.

FAQ

Can I search with more than one keyword set in one run?

Yes. The queries input accepts an array, and each query is scraped independently.

What field tells me where the image came from?

Use sourcePageUrl for the page and sourceName for the hosting website.

Which field is the direct image file?

Use imageUrl. The thumbnailUrl is the small Google preview image.

How do I get more than one page of results?

Increase maxResultsPerQuery. The Actor paginates automatically up to 500 results per query.

Can I use this for image provenance checks?

Yes. The returned sourcePageUrl, sourceName, and imageUrl fields support provenance review in downstream workflows.

Related Actors

• Use Google Maps Reviews Scraper - useful after this Actor when a returned Google record points to a business or place and the next step needs reviews.
• Use Google Ads Scraper - Live Search Sponsored Ads - a sibling workflow for Google ad research when the next step is sponsored search results.
• Use Google Ads Transparency Scraper - Creatives & Run Dates - a sibling workflow for ad creative and run-date research.

Responsible use

Use the returned image links and source pages in line with the source site's terms, Google's terms, and applicable copyright and licensing rules. The usageRights filter reflects Google's labeling and is useful for discovery, while the source page remains the place to verify the actual license before reuse or redistribution.