Yandex Reverse Image Search API

Search by image instead of by keywords. Give this API the URL of any image and it searches Yandex's reverse image engine (the same one behind yandex.com/images) to find every page where the image appears online, visually similar images, other sizes and resolutions of the same image, products that match the image, and descriptive tags. Pay per result, with a hard cap you control.

Yandex is widely regarded as one of the strongest reverse image engines on the web, especially for faces, places, and content that Western engines miss. This actor gives you that power as a clean JSON API.

What this API returns

Each result is one dataset row, tagged with a result_type field so you can filter:

result_type	What it is	On by default
matching_page	A page where the image (or a close match) appears online: title, link, snippet, source	✅
similar_image	A visually similar image from across the web: thumbnail, link	✅
image_size	Another resolution of the same image, categorized large / medium / small	❌
image_tag	A suggested search term describing what is in the image	❌
shopping_result	An e-commerce product matching the image, with price where available	❌
knowledge_graph	Entity card when the image contains a recognizable person, place, or brand	❌

Every row also carries the query image URL, the Yandex domain used, the crop box (if any), and a search timestamp.

Use cases

• Find where your images are used: track unauthorized use of product photos, artwork, or brand assets across the web.
• Brand protection and counterfeit detection: enable shopping matches to find marketplaces selling products that look like yours.
• Image provenance and fact-checking: find earlier and higher-resolution versions of an image to trace where it first appeared.
• Face and person search research: Yandex's reverse image engine is known for strong face matching (use responsibly and lawfully).
• Duplicate and similar image discovery: build datasets of visually similar images for ML or content curation.
• Source hunting: find the original, full-resolution version of a compressed or cropped image.

Quick start

The fastest possible run, just paste an image URL:

{
  "image_url": "https://substack-post-media.s3.amazonaws.com/public/images/edbfb2cd-ebcb-4527-bec7-5315c182278f_445x445.png"
}

Search only part of the image with a crop box (fractions of width and height, left;top;right;bottom):

{
  "image_url": "https://example.com/group-photo.jpg",
  "crop": "0.1;0.2;0.5;0.8",
  "include_similar_images": true
}

Everything on, capped at 100 rows:

{
  "image_url": "https://example.com/product.jpg",
  "include_matching_pages": true,
  "include_similar_images": true,
  "include_image_sizes": true,
  "include_image_tags": true,
  "include_shopping_results": true,
  "include_knowledge_graph": true,
  "max_results": 100
}

Input parameters

Parameter	Type	Required	Default	Description
image_url	string	✅	-	Public http(s) URL of the image to search by. Yandex fetches it, so it must be reachable from the internet (direct image link, no login or redirect).
crop	string	❌	(none)	Crop the image before searching: four ;-separated fractions 0-1 in the order left;top;right;bottom, e.g. 0.1;0.2;0.9;0.8.
include_matching_pages	boolean	❌	true	Pages where the image appears online (result_type matching_page).
include_similar_images	boolean	❌	true	Visually similar images (result_type similar_image).
include_image_sizes	boolean	❌	false	Other resolutions of the same image (result_type image_size).
include_image_tags	boolean	❌	false	Descriptive tags for the image content (result_type image_tag).
include_shopping_results	boolean	❌	false	Matching products with prices (result_type shopping_result).
include_knowledge_graph	boolean	❌	false	Entity card for recognizable subjects (result_type knowledge_graph).
yandex_domain	string	❌	"yandex.com"	Regional domain: yandex.com, yandex.ru, yandex.by, yandex.kz, yandex.uz, yandex.com.tr.
max_results	integer	❌	0	Hard cap on rows returned and billed. 0 = everything found. Set 20 for a cheap first run.

Example output row

{
  "result_type": "matching_page",
  "position": 27,
  "title": "Central Illustration Agency Illustration portfolio: Matt Taylor",
  "link": "https://tr.pinterest.com/pin/matt-taylor-digital-illustration-illustrator-graphic-poster-art-car-scenery-bold--826762444071832079/",
  "thumbnail": "https://avatars.mds.yandex.net/i?id=592d913ed53e1f78fdd0804bd7064417b709bafb-5437458-images-thumbs&n=13&w=296&h=180",
  "original": "https://i.pinimg.com/736x/9e/b1/f7/9eb1f76ee45eb01706be89b8748a911b.jpg",
  "source": "tr.pinterest.com",
  "snippet": "After graduating from Buckinghamshire University he rolled straight into a successful ten year illustration career.",
  "image_url": "https://substack-post-media.s3.amazonaws.com/public/images/edbfb2cd-ebcb-4527-bec7-5315c182278f_445x445.png",
  "yandex_domain": "yandex.com",
  "crop": "",
  "search_timestamp": "2026-07-05T12:00:00"
}

Two dataset views are available in Apify Console and via API: Overview (all rows, typed) and Visual results (thumbnail-forward).

Pricing

This actor uses pay-per-event billing: you are charged one result_returned event per dataset row. You control cost two ways:

1. Result-type toggles: only the sections you enable produce rows. The default (matching pages + similar images) covers the two most common reverse image questions.
2. max_results cap: a hard ceiling on rows returned and billed per run.

A search that legitimately finds nothing still pushes one no_results summary row so billing stays transparent. Invalid input and upstream failures do not produce billable rows.

The current per-result price is shown on the Apify Store listing. New users get Apify's free trial credit, which covers many test runs.

How to get started

1. View on Apify Store and click Try for free.
2. Paste an image URL into the input form and run.
3. Read the results from the dataset (JSON, CSV, or Excel), or call the actor from the Apify API in your own code.

🔌 Use this API from Claude (MCP)

Add this actor as a tool in Claude Code (free trial), Claude Cowork (free trial), or any other MCP client, via the hosted Apify MCP server. Use this actor-specific URL:

https://mcp.apify.com/?tools=actors,docs,johnvc/yandex-reverse-image-search

Setup walkthrough:

Apify MCP integration docs: https://docs.apify.com/platform/integrations/mcp

FAQ

How is billing calculated?

One dataset row = one billed result. Enable only the result types you need and set max_results to cap any run. An empty search pushes a single summary row; failed runs bill nothing.

What images work best?

Publicly reachable, reasonably sized images (300px+ on the short side). Yandex fetches the URL you provide, so private, localhost, or login-protected URLs will not work. If you only care about part of an image (one face in a group photo, one product on a shelf), use the crop parameter.

How is this different from a normal image search?

A normal image search takes keywords and returns images. Reverse image search takes an image and returns where it appears, what looks like it, and what it contains. For keyword-based Yandex image search, see the Yandex Search Scraper API, which also covers web results, ads, and the knowledge graph.

Which Yandex domain should I use?

yandex.com (the default) works for most cases. Regional domains like yandex.ru can surface more results for images that circulate mainly on the Russian-language web.

Is it legal to reverse image search a person?

Laws differ by jurisdiction, and platform terms apply. This tool returns publicly indexed data only; you are responsible for using it lawfully and ethically, especially for images of people.

Featured Tasks

Ready-to-run examples that show this API solving a specific problem. Each opens its own setup so you can run it on your account in one click.

• Reverse image search in Claude via MCP - let Claude reverse-search any image through Yandex live in the conversation via MCP.
• Find where an image appears online with Yandex - every page where an image appears, with source and snippet, as JSON.
• Trace an image to its original source with Yandex - find an image's original source and every page using it.
• Find visually similar images from an image URL - visually similar images from across the web, with thumbnail, full-size URL, and source.
• Find products from an image with Yandex - visually similar products and shop listings from a product image.
• Reverse image search with Yandex, results as JSON - matching pages, similar images, and more, as structured JSON, no API key setup.

Last Updated: 2026.07.10