Bulk Image Downloader & Google Images Scraper

Bulk Image Downloader & Google Images Scraper to ZIP

Download images in bulk: search Google Images with every filter (size, color, type, license, date), paste a list of image URLs, or take the image fields of any dataset (Amazon, Shopify, Instagram, real estate scrapers). You get the full-size files in a ZIP with a metadata.csv, each file also with a direct link, plus a table of image URL, size, rank and source page.

Built for AI and computer-vision datasets, product catalogs, mood boards and content work: no duplicates, minimum resolution checked before download, convert WebP/AVIF to JPEG, resize to 512 or 1024 px. Scheduled runs download only new images.

No browser, no API key, no proxy setup.

Why this Actor

📦 Files, not just links	Other Google Images scrapers return URLs and leave the downloading to you. Here every image is downloaded, checked and saved: one ZIP (in ~200 MB parts) with metadata.csv, and each file in the key-value store with a link
🔎 Three sources, one run	Google Images searches, a list of image URLs, or another Actor's dataset (image fields are found automatically)
🧬 No duplicates	The same picture resized or recompressed on different sites is downloaded once (perceptual hash)
📏 Pay only for images you want	Minimum width/height is checked against Google's data before downloading; failed downloads, duplicates and small images are free
🤖 AI-ready	Convert to JPEG/PNG/WebP, shrink the longest side to 512/1024 px, keep ranks and source pages for labeling
🛟 Every image saved	Downloads with a real browser fingerprint; when a site blocks the original (Instagram, Facebook), Google's thumbnail is saved instead and marked
🔔 Only new images	Scheduled runs skip everything already downloaded, including the same picture at a new URL

How it works

1. Add Google Images searches (e.g. red fox), and/or Image URLs, and/or Datasets with image links
2. Optional: Google filters (Size, Color, Type, Usage rights, Time, Only from this site), Minimum width/height, Convert to, Resize
3. Run. Click ZIP archive in the Output tab to download all images, or click the file link of a row to save one image

How to download the images

Way	How
All images at once	Output tab → ZIP archive: images.zip downloads with every file in folders and metadata.csv. The run's status message has the same download link
One image	Click the file link (fileUrl) in the table row: the file is saved to your computer
Every file separately	Storage tab → Key-value store: each image is a record you can download
From code	fileUrl of each row is a direct download link; the ZIP is the record images.zip of the run's key-value store

What you get

A cloud run in September 2026: 3 searches (eiffel tower, golden retriever puppy, modern kitchen interior), 200 images each.

Original files saved	559 of 600 (93%)
With the thumbnail fallback	600 of 600
Time	~2 minutes for 600 files (199 MB ZIP)
Formats	JPEG 515, WebP 59, PNG 25, GIF 1 (originals as the site serves them)
Images per search	~240 unique (Google shows ~250 per search; use more specific searches for more)
Image size from Google vs. the real file	equal in 331 of 331 checked

Output

Field	Description
fileUrl	Direct link to the saved file
fileName, zipFile	Path of the file inside the ZIP (red-fox/005_a-z-animals.com.jpg) and which ZIP part
fileWidth, fileHeight, fileSizeKB, format	The saved file (after conversion or resizing)
imageUrl, pageUrl, domain	The original image and the page it is on
width, height	Size of the original, from Google
query, position	The search and the rank in Google Images (or the URL / dataset the image came from)
thumbnailUrl, imageId	Google thumbnail and Google's image ID
sourceField, sourceTitle	Dataset source only: the field the link was in and the item's title or name
downloadedFrom	original or thumbnail
downloaded, error	Failed downloads are listed with the reason (free)
hash	Perceptual hash (dHash), for your own de-duplication

images.zip contains the files in one folder per search and a metadata.csv with the same fields.

Google Images filters

Filter	Options
Size	Large, Medium, Icon, Larger than 2/4/6/8/10/12/15/20/40/70 MP
Color	Full color, Black and white, Transparent, 12 colors
Type	Photo, Face, Clip art, Line drawing, Animated (GIF)
Usage rights	Creative Commons licenses, Commercial & other licenses
Time	Past 24 hours, week, month, year
Aspect ratio	Tall, Square, Wide, Panoramic
File type	JPG, PNG, GIF, WebP, SVG, BMP, ICO
Site	Only images from one website
Country, language, SafeSearch	Any Google country and language

Images from another Actor's dataset

Put the dataset ID (or its link from Console) into Datasets with image links. Image links are found in fields such as image, images, thumbnail, photos, logo, or any link ending in .jpg, .png, .webp. To take specific fields only, list them in Image fields, dotted for nested ones: images, product.mainImage, photos.url.

Each row keeps the item's title and link, so product photos stay matched to their products.

Only new images: monitoring

1. Switch on Only new images and set a Monitor name (one per task)
2. Save as a task and schedule it (daily or weekly)
3. The first run downloads and remembers the images. Later runs download only images that are new, and skip pictures you already have even at a different URL

Pricing

Pay per image, no subscription:

Event	Price
Run start	$0.002
Image downloaded (file saved + row)	$0.002
Image result without download (Download image files off)	$0.0005

Free: failed downloads, duplicates, images below your minimum size, already-seen images in monitoring.

Examples: 1,000 images downloaded = $2. A table of 1,000 Google Images results without files = $0.50.

Ready-made tasks

Open one, change the search, and run:

Task	What you get
Download Google Images in Bulk: Full-Size Files in a ZIP	All images of a search as full-size files in one ZIP
Bulk Image Downloader from a List of URLs	Your image links downloaded, converted to JPG
AI Image Dataset Builder: Google Images to 512px JPEG	One folder per class, 512 px JPEG, no duplicates

Input example

{
    "queries": ["golden retriever", "labrador retriever", "german shepherd"],
    "maxImagesPerQuery": 200,
    "imageSize": "large",
    "minWidth": 800,
    "convertTo": "jpeg",
    "maxDimension": 512
}

Tips and limits

• About 250 images per search is Google's own limit. For more, split a broad search into specific ones (red fox winter, red fox cub, red fox running).
• Minimum size for Google searches is checked before downloading; for URL lists and datasets it is checked after (still free).
• Usage rights: Google's filter shows how an image is labeled; check the license on the source page before commercial use.
• Big runs: files are packed into ZIP parts of ~200 MB (images.zip, images-2.zip...). Turn off Save each file if you only need the ZIP.
• Titles: Google's lightweight results page this Actor reads has no image titles; use domain and pageUrl.