Yandex Reverse Image Search MCP server

Registry name: io.github.johnisanerd/yandex-reverse-image-search. Maintained by AlphaOSINT (Apify developer johnvc, GitHub johnisanerd).

What it does

Search Yandex by image URL and get the pages where the image appears, visually similar images, other sizes, matching shop products and image tags.

Searches by image rather than by text. Use it for image provenance, copyright checks, product matching and OSINT investigations.

This server is not self-hosted. It runs on Apify’s hosted MCP endpoint (mcp.apify.com), which starts the Apify Actor johnvc/yandex-reverse-image-search for each tool call. You need an Apify account, and every call is billed to that account at the Actor’s prices below. Apify’s own pricing and terms apply to the account.

Charge Price (USD) What it covers
Result returned $0.0065 One reverse image result row: a page where the image appears, a similar image, an alternate size, an image tag, a shopping match, or a knowledge graph card.

Prices shown are Apify’s free-plan rates; paid Apify plans get lower per-event prices. Apify also bills a token amount per run start and per stored row ($0.00001, $0.00001). Prices change from time to time; the Actor’s page on Apify shows the current ones.

Data source and terms

Yandex’s public reverse image search results, collected by scraping. The server returns the data as published at the time of the call. You are responsible for using it in line with the source’s terms and the laws that apply to you. AlphaOSINT is not affiliated with the source.

Endpoint

Setup

Claude Code

claude mcp add --transport http yandex-reverse-image-search "https://mcp.apify.com/?tools=johnvc/yandex-reverse-image-search"

VS Code (.vscode/mcp.json)

{
  "servers": {
    "yandex-reverse-image-search": {
      "type": "http",
      "url": "https://mcp.apify.com/?tools=johnvc/yandex-reverse-image-search"
    }
  }
}

Cursor (~/.cursor/mcp.json)

{
  "mcpServers": {
    "yandex-reverse-image-search": {
      "url": "https://mcp.apify.com/?tools=johnvc/yandex-reverse-image-search"
    }
  }
}

Claude Desktop (Settings → Connectors → Add custom connector, paste the URL). Older versions without remote connectors can bridge through mcp-remote in claude_desktop_config.json:

{
  "mcpServers": {
    "yandex-reverse-image-search": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-remote",
        "https://mcp.apify.com/?tools=johnvc/yandex-reverse-image-search"
      ]
    }
  }
}

Each client opens an Apify sign-in (OAuth) on first use. To use a token instead, send Authorization: Bearer <APIFY_TOKEN> as a header; create one under Apify Console → Settings → API & Integrations.

Tools

Tool What it does
johnvc--yandex-reverse-image-search Runs the Actor with the parameters below and returns the result rows.
get-actor-run Status of a run started by the tool, for long runs.
get-dataset-items Pages through a run’s full output by datasetId when it is larger than the inline result.
get-key-value-store-record Reads a record a run stored, such as a run report.
abort-actor-run Stops a run.

Apify’s endpoint adds the four helper tools automatically whenever an Actor tool is loaded.

Parameters of johnvc--yandex-reverse-image-search

| Parameter | Type | Required | Default | Description | |—|—|—|—|—| | image_url | string | yes | "https://substack-post-media.s3.amazonaws.com/public/images… | Provide the public http(s) URL of the image to search by. Yandex fetches this URL, so it must be reachable from the internet (no localhost or private links). Example: https://substack-post-media.s3.amazonaws.com/public… | | crop | string | | | Optionally crop the image before searching, as four ';'-separated fractions between 0 and 1 in the order left;top;right;bottom. Example: 0.1;0.2;0.9;0.8 searches only the middle of the image. Leave blank to search the… | | include_matching_pages | boolean | | true | Return pages where this image (or a close match) appears online, with page title, link, and snippet. Rows with result_type 'matching_page'. On by default. | | include_similar_images | boolean | | true | Return visually similar images from across the web. Each row carries a thumbnail, the full-size image URL, and the site hosting it. Yandex returns no page title or snippet for this result type, so those columns stay emp… | | include_image_sizes | boolean | | false | Return other resolutions of the same image, categorized large / medium / small. Rows with result_type 'image_size'. Off by default. | | include_image_tags | boolean | | false | Return suggested search terms describing what is in the image. Rows with result_type 'image_tag'. Off by default. | | include_shopping_results | boolean | | false | Return e-commerce products that match the image content, with prices where available. Yandex only fills this section for a minority of images, so many searches return no product rows at all. Rows with result_type 'shopp… | | include_knowledge_graph | boolean | | false | Return the entity card when the image contains a recognizable subject (a person, landmark, product, or brand). One row with result_type 'knowledge_graph'. Off by default. | | yandex_domain | string: yandex.com, yandex.ru, yandex.by, yandex.kz, yandex.uz, yandex.com.tr | | “yandex.com” | The regional Yandex domain to search from. Defaults to 'yandex.com'. Results can differ by region. | | max_results | integer | | 50` | Cap the total number of result rows returned by this run. Results are sold in blocks of 10: the minimum is 10, and any other value rounds up to the next multiple of 10 (15 becomes 20). Every search bills at least one bl… |

Example arguments:

{
  "image_url": "https://substack-post-media.s3.amazonaws.com/public/images/edbfb2cd-ebcb-4527-bec7-5315c182278f_445x445.png",
  "max_results": 50
}

Output

The tool returns JSON rows with these fields.

Field Type Description
result_type string: matching_page, similar_image, image_size, image_tag… Discriminator for the row kind: ‘matching_page’ (a page where the image appears), ‘similar_image’, ‘image_size’ (another resolution of the same image), ‘image_tag’ (descriptive search term), ‘shopping_result’, ‘knowledg…
position integer 1-based rank of this result within its result type.
title string Title of the page, image, product, or entity. Populated for ‘matching_page’, ‘image_tag’, ‘shopping_result’ and ‘knowledge_graph’ rows. Yandex returns no title for ‘similar_image’ or ‘image_size’ rows, so it is empty th…
link string URL of the page, image, or product this row points at.
thumbnail string Thumbnail image URL for this result.
original string Full-size image URL when the result carries one. Populated for ‘matching_page’ rows, and for ‘similar_image’ and ‘image_size’ rows where it is read back from the result link.
source string Source site or platform name for this result. Populated for ‘matching_page’ rows from the page’s own site name, and derived from the image host for ‘similar_image’ and ‘image_size’ rows.
source_link string URL of the source site when the source was returned as an object.
snippet string Text snippet from the page where the image appears. Populated for ‘matching_page’ rows only. Yandex returns no snippet for any other result type.
thumbnail_width integer Width in pixels of the thumbnail image, when Yandex reports it.
thumbnail_height integer Height in pixels of the thumbnail image, when Yandex reports it.
original_width integer Width in pixels of the full-size image, when Yandex reports it. Useful for picking the highest-resolution copy of an image.
original_height integer Height in pixels of the full-size image, when Yandex reports it.
description string For ‘knowledge_graph’ rows: the entity blurb Yandex returns for a recognizable subject.
text string For ‘image_tag’ rows: the suggested search term describing the image. Also copied into ‘title’.
size string For ‘image_size’ rows: the pixel dimensions of this alternate resolution, as reported by Yandex.
size_category string For ‘image_size’ rows: which size bucket this resolution belongs to (large, medium, or small).
image_url string The input image URL this reverse search was run for.
yandex_domain string The regional Yandex domain the search ran on.
crop string The crop box applied before searching, if any (left;top;right;bottom fractions).
search_timestamp string ISO 8601 timestamp of when the search was executed.
error_message string For ‘error’ rows: a human-readable explanation of what went wrong.