Group visually similar images

What this example returns

The dataset’s “Duplicates and matches” view shows 15 columns: #, Image, Verdict, Group, Keep, Duplicate of, Compared with, Similarity, Hash distance, Near duplicate, Similar, Width, Height, Custom ID, and Error. See every output field.

What each setting does

The input this example runs Duplicate Image Finder & Image Similarity with. “Try this input” opens it in the console, where you can run it as-is or change any field first.

Images imageUrls

https://upload.wikimedia.org/wikipedia/commons/thumb/a/a8/Tour_Eiffel_Wikimedia_Commons.jpg/500px-Tour_Eiffel_Wikimedia_Commons.jpg, https://upload.wikimedia.org/wikipedia/commons/thumb/8/85/Tour_Eiffel_Wikimedia_Commons_%28cropped%29.jpg/500px-Tour_Eiffel_Wikimedia_Commons_%28cropped%29.jpg, https…

Links to the images, one per line: JPEG, PNG, WebP, GIF, AVIF, HEIC, TIFF or BMP, up to 100 MB each. With no reference image, every image is checked against all the others and duplicates are grouped. With a reference image, every image is compared with it. Google Drive, Dropbox and GitHub share links are turned into…

What to compare comparisonMode

Visual similarity only

Both: a visual similarity score from a vision model plus perceptual hashes that prove whether two files are the same picture (resized, recompressed or lightly edited). Hashes only: duplicate detection without the model, the fastest. Visual similarity only: the model score without hashes.

Default: Visual similarity + duplicate hashes → this example uses Visual similarity only

Similarity threshold (0 to 1) similarityThreshold

0.85

Two images count as similar ('isSimilar') when the vision model's similarity score is at least this value. 0.85 suits most photos; raise it to 0.95 for near-identical scenes only, lower it to 0.75 for loosely related images.

Group similar images too groupSimilar

Yes

Off: groups hold duplicates only (the same picture). On: images whose similarity score reaches the threshold join the same group too, for example several shots of one product. Only used when no reference image is set.

Default: No → this example uses Yes

{
  "imageUrls": [
    "https://upload.wikimedia.org/wikipedia/commons/thumb/a/a8/Tour_Eiffel_Wikimedia_Commons.jpg/500px-Tour_Eiffel_Wikimedia_Commons.jpg",
    "https://upload.wikimedia.org/wikipedia/commons/thumb/8/85/Tour_Eiffel_Wikimedia_Commons_%28cropped%29.jpg/500px-Tour_Eiffel_Wikimedia_Commons_%28cropped%29.jpg",
    "https://upload.wikimedia.org/wikipedia/commons/thumb/0/0c/GoldenGateBridge-001.jpg/500px-GoldenGateBridge-001.jpg",
    "https://upload.wikimedia.org/wikipedia/commons/thumb/b/bf/Golden_Gate_Bridge_as_seen_from_Battery_East.jpg/500px-Golden_Gate_Bridge_as_seen_from_Battery_East.jpg",
    "https://upload.wikimedia.org/wikipedia/commons/thumb/3/3a/Cat03.jpg/960px-Cat03.jpg",
    "https://upload.wikimedia.org/wikipedia/commons/thumb/9/90/Labrador_Retriever_portrait.jpg/960px-Labrador_Retriever_portrait.jpg"
  ],
  "comparisonMode": "semantic",
  "groupSimilar": true,
  "similarityThreshold": 0.85
}

What it costs

You pay for the events the actor reports, such as each result or page it produces. Platform usage is included in the event prices: compute, proxies, browser and storage cost nothing extra.

• Image resultfrom $0.00034One result row: an image checked against the other images of the list (its duplicate group, closest match and scores), or one image compared with one reference image. Images that cannot be read get free error rows.

Full pricing

When to use this example

From the Duplicate Image Finder & Image Similarity readme:

Other ways to run Duplicate Image Finder & Image Similarity

• Find duplicate photos in a listGroups resized copies of the same photo and marks the largest copy to keep.
• Compare images with reference photosScores every image against your originals and flags copies.
• Fast duplicate check (hashes only)Hashes only, the fastest mode; broken links get free error rows.

All examples of Duplicate Image Finder & Image Similarity

About Duplicate Image Finder & Image Similarity

Find duplicate and near-duplicate images in a list of links or a dataset, or compare images with your reference images. Perceptual hashes (pHash, dHash) catch resized and recompressed copies; a vision model scores visual similarity. Duplicate groups, the copy to keep, and a verdict per image.

Price:

from $0.34 / 1,000 results

Developer:

Echofield

Categories:

Developer tools, AI, E-commerce

Duplicate Image Finder & Image Similarity pageOutput fieldsPricing