Google Images Scraper - Image URLs, Sizes & Sources

Google Images Scraper extracts image search results from Google Images: the original image URL, its width and height, the page it is on, that page's title, the domain, the file type and Google's thumbnail. Paste search terms or Google Images URLs, use any of Google's filters (size, color, type, file type, shape, usage rights, date), and get one row per image. No login, no API key, no browser, from $0.70 per 1,000 images.

{
  "query": "golden retriever puppy",
  "position": 1,
  "imageUrl": "https://blog.ollie.com/wp-content/uploads/250506_Livingroom_SHOT-10_0173_FR-scaled.jpg",
  "imageWidth": 2560,
  "imageHeight": 1707,
  "title": "Golden Retriever: Pros, Cons & Diet Guide | Ollie",
  "pageUrl": "https://blog.ollie.com/breed-deets-golden-retriever/",
  "domain": "blog.ollie.com",
  "fileType": "jpg",
  "thumbnailUrl": "https://encrypted-tbn0.gstatic.com/images?q=tbn:ANd9GcRbrtNwKnE9g9GMxWrwJRdj6a2qZ-ewk6Ri2OYZ2NLl7GsX32lUZok3vVJK&s",
  "megapixels": 4.37,
  "orientation": "landscape"
}

Benchmark: 4,195 images with page titles from 10 searches in 3 minutes 9 seconds at 256 MB, every request answered (measured on Apify, 2026-10-29).

What does Google Images Scraper do?

It searches Google Images the way a person does and turns the results into clean JSON, CSV or Excel. Give it keywords, Google operators or ready-made Google Images URLs, pick a country and language, narrow the results with Google's own filters, and it returns every image with the page it comes from. It is built for image datasets and AI training data, brand and product image monitoring, visual SEO research, content sourcing and as an image search tool for AI agents.

• Any search, up to Google's limit. Search terms, operators (site:, -word, "exact phrase") or pasted Google Images URLs with their filters. Up to 600 images per search, which is as many as Google shows.
• The original, not just a thumbnail. Every row has the full-size image URL on the website that hosts it, its pixel size, megapixels and orientation, plus Google's always-loadable thumbnail.
• Where each image comes from. The source page's URL, its domain and its title, so you know the context and who to ask about the licence.
• Every Google Images filter. Size (large, medium, icon, larger than 2-70 MP), color (full color, black and white, transparent, 12 single colors), type (photo, face, clip art, line drawing, animated), file type (JPG, PNG, GIF, WEBP, SVG, BMP, ICO), shape, Creative Commons or commercial licences, past day/week/month/year or a custom date range, SafeSearch, any country and interface language.
• Only pay for images. Searches with no images and searches Google refused cost nothing, and duplicates within a search are removed before you pay.
• Drop-in for hooli/google-images-scraper. The same input names (queries, maxResultsPerQuery) and the same output fields (query, title, imageUrl, imageWidth, imageHeight, thumbnailUrl, thumbnailWidth, thumbnailHeight, origin, contentUrl), so an existing integration works by changing the Actor ID.
• Built for automation. Schedules, the Apify API, webhooks, Make, Zapier, n8n, Google Sheets, and AI agents through Apify's MCP server.

Why it is fast and cheap

Google Images in a desktop browser is a JavaScript application that scrapers usually drive with a headless browser or a paid SERP proxy. This Actor asks Google for the lightweight page it serves to simple phones instead: 20 images in about 23 KB, with the original image URL, source page and sizes in plain links, plus an 8 KB detail page per image for its page title. No browser, no paid proxy, so a search of 100 images takes a few seconds and the price stays low.

What data can you extract from Google Images?

Field	Example	What it is
imageUrl	https://blog.ollie.com/.../SHOT-10_0173_FR-scaled.jpg	The full-size original image
imageWidth, imageHeight	2560, 1707	Size of the original in pixels
title	Golden Retriever: Pros, Cons & Diet Guide | Ollie	Title of the page the image is on
pageUrl	https://blog.ollie.com/breed-deets-golden-retriever/	The page the image is on
domain	blog.ollie.com	That page's site
fileType	jpg	Image format
thumbnailUrl	https://encrypted-tbn0.gstatic.com/images?q=tbn:...	Google's thumbnail, always loadable
thumbnailWidth, thumbnailHeight	275, 183	Thumbnail size
megapixels	4.37	Width x height in millions of pixels
orientation	landscape	landscape, portrait or square
position	1	Rank in Google Images for the search, duplicates removed
page	1	Which results page (20 images each) it was on
query	golden retriever puppy	The words searched
correctedQuery	mountain landscape	What Google searched instead when it autocorrected; else null
imageId	eYoKU9jW-eRT6M	Google's ID for the result
searchUrl	https://www.google.com/search?q=...&tbm=isch...	The search, to open in a browser
input	golden retriever puppy	Your search line exactly as entered
scrapedAt	2026-10-29T12:00:00.000Z	When this run collected the row

How to scrape Google Images

1. Click Try for free and type your searches into Search queries, one per line. A Google Images URL works too.
2. Set Max images per query (100 by default, up to 600). You pay per image, so this is also your cost cap.
3. Pick filters, country and language if you need them, then click Start. 100 images take about 10 seconds, 1,000 about a minute.
4. Open the Output tab, or download the results as JSON, CSV, Excel or HTML.

How much does it cost to scrape Google Images?

You pay per image returned, and the price drops on higher Apify plans:

Apify plan	Per 1,000 images
Free	$1.20
Starter	$1.00
Scale	$0.85
Business	$0.70

• 10 product searches of 100 images each (1,000 images) cost $1.00 on the Starter plan.
• 50 brand searches of 600 images each (30,000 images) cost $21.00 on the Business plan.
• A run also costs $0.00005 to start. A search that fails, or finds nothing, costs nothing more. Page titles are included in the price.
• The Free plan's $5 monthly credit covers about 4,000 images.

Input

{
  "queries": ["golden retriever puppy", "https://www.google.com/search?q=coffee+cup&tbm=isch&tbs=ic:trans"],
  "maxResultsPerQuery": 100,
  "imageSize": "large",
  "usageRights": "creativeCommons",
  "countryCode": "us"
}

Field	Default	Notes
Search queries (queries)	golden retriever puppy	Search terms or Google Images URLs, one per line
Max images per query (maxResultsPerQuery)	100	Up to 600, Google's own limit
Max images in total (maxItems)	10,000	Cap for the whole run
Size, Color, Type, File type, Shape, Usage rights, Time	any	Google's Tools filters; a pasted URL keeps its own
From date / To date (dateFrom, dateTo)	none	Custom date range, YYYY-MM-DD
SafeSearch (safeSearch)	on	Off returns unfiltered results
Country (countryCode), Interface language (languageCode)	United States, the country's language	Google's gl and hl
Search exactly as typed (exactQuery)	off	Turns off Google's autocorrect
Page titles and file types (includePageDetails)	on	One small extra request per image, included in the price
Proxy	Apify Proxy (datacenter)	Switches to it only when Google rate-limits the run's own address; Residential is used only if you choose it, and only for results pages

Output

Each image is one row in the dataset. The Overview view shows the image, page title, links and size; Size and format and Search and errors show the rest, and every export includes all fields.

{
  "recordType": "image",
  "query": "golden retriever puppy",
  "position": 1,
  "imageUrl": "https://blog.ollie.com/wp-content/uploads/250506_Livingroom_SHOT-10_0173_FR-scaled.jpg",
  "imageWidth": 2560,
  "imageHeight": 1707,
  "title": "Golden Retriever: Pros, Cons & Diet Guide | Ollie",
  "pageUrl": "https://blog.ollie.com/breed-deets-golden-retriever/",
  "domain": "blog.ollie.com",
  "fileType": "jpg",
  "thumbnailUrl": "https://encrypted-tbn0.gstatic.com/images?q=tbn:ANd9GcRbrtNwKnE9g9GMxWrwJRdj6a2qZ-ewk6Ri2OYZ2NLl7GsX32lUZok3vVJK&s",
  "thumbnailWidth": 275,
  "thumbnailHeight": 183,
  "megapixels": 4.37,
  "orientation": "landscape",
  "imageId": "eYoKU9jW-eRT6M",
  "page": 1,
  "correctedQuery": null,
  "searchUrl": "https://www.google.com/search?q=golden+retriever+puppy&tbm=isch&hl=en&gl=us&ie=UTF-8&oe=UTF-8&safe=active",
  "input": "golden retriever puppy",
  "scrapedAt": "2026-10-29T11:55:27.146Z",
  "origin": "blog.ollie.com",
  "contentUrl": "https://blog.ollie.com/breed-deets-golden-retriever/"
}

A search that cannot be read becomes a row with "recordType": "error" and an errorCode instead of failing the whole run:

errorCode	Meaning
RATE_LIMITED	Google kept refusing the run's addresses; run again later, or choose Residential in the proxy settings
BLOCKED	Google refused the request; same remedies
NETWORK	Google could not be reached
PARSE_FAILED	Google answered with a page this Actor does not know; please report it

Error rows are never charged. Searches with no images at all are listed in the log and in the run summary.

Tips and limits

• Google shows at most about 400-600 images for one search. To collect more, split the topic into narrower searches ("red sneakers men", "red sneakers women") or run the same search with different filters or countries.
• Use operators to focus. site:unsplash.com mountains returns images from one site only; -stock drops pages that mention "stock".
• Turn off page titles for the fastest runs. Each image then costs one request less; the file type then comes from the image URL and is missing for URLs without an extension.
• Images are not downloaded. You get their URLs; the originals stay on the websites that host them, and a few sites block hotlinking. The thumbnail URL always loads.
• Licences. The usage-rights filter is Google's and relies on the licence information pages publish. Check the licence on the source page before you reuse an image.
• Exact sizes are not offered because Google ignores them on this page; use "Larger than" sizes instead.
• Hit your spending limit? The run stops cleanly at your maximum cost per run and keeps everything it collected. Raise the limit in the run options to get more.

FAQ

Is it legal to scrape Google Images?

This Actor collects only what Google Images shows publicly, without logging in, and does not download the images themselves. The images belong to their owners and may be protected by copyright; results can also contain personal data, which laws such as the GDPR protect. Use the data for a legitimate purpose, check licences before reusing images, and ask a lawyer if you are unsure. Google and Google Images are trademarks of Google LLC; this Actor is not affiliated with or endorsed by Google.

Can I use it through the API or with Make, Zapier and n8n?

Yes. Every run can be started through the Apify API and returns the dataset as JSON, CSV or Excel. Schedules, webhooks and integrations for Make, Zapier, n8n and Google Sheets are on the Integrations tab.

Can AI agents use it?

Yes. Agents connected to Apify's MCP server can find and run this Actor as a tool, with the same per-image pricing. A call with just {"queries": ["..."]} works; every other field has a sensible default.

I use hooli/google-images-scraper. Will my integration work?

Yes. Send the same input (queries, maxResultsPerQuery) and read the same fields (query, title, imageUrl, imageWidth, imageHeight, thumbnailUrl, thumbnailWidth, thumbnailHeight, origin, contentUrl). The rows also carry the extra fields listed above.

Why did I get fewer images than I asked for?

Google has fewer images for that search and filter combination than your limit (it never shows more than about 600), duplicates were removed, the run reached your maximum cost per run, or it reached its timeout. The log says which, and nothing collected before that is lost.

Related scrapers

• Google Search Results Scraper: the web, News and Videos tabs of Google, with organic results, places and top stories.
• Pinterest Scraper: images and pins from Pinterest, the other big place people search for pictures.
• Website Brand Extractor: logos, colors and brand assets from the websites your images come from.

Support

Missing a field, or found a search that does not work? Open an issue on the Issues tab with the input you used, and it will be looked at quickly.