Adobe Stock Scraper: Assets, Keywords & Portfolio Changes

For contributors, creative research teams and catalog analysts who want an exportable record of Adobe Stock search results and contributor portfolios. Collect asset-specific keywords, titles, public creator attribution, dimensions, AI flags, preview links and the license information Adobe supplies. About $0.0005 per delivered asset, plus $0.0023 per allocated GB to start (minimum one unit). At the default 1 GB, 20 assets cost $0.0123 and 1,000 assets cost $0.5023 at the base rate. No Adobe login or Adobe API key is required.

Use the same input repeatedly to observe which asset IDs appear in your selected search or portfolio. Turn on compareWithPreviousRun to add a previous observation timestamp to assets already seen in a completed run. Export the unique-asset table as JSON, CSV or Excel.

What you get

Each row identifies a public Adobe Stock asset. Search suggestions are never presented as an asset's keywords. The keyword list comes from the individual asset's public metadata. Connection sessions can be reused anonymously; asset metadata is fetched again on each run. Repeated appearances across your selected searches or portfolios produce one delivered asset and one asset charge in that run.

Field	Meaning
verdict	A readable media label and asset title.
recordType	ASSET for useful data; NOTE for a free explanation.
assetId	Adobe Stock asset identifier, stored as a string.
title	Title supplied by Adobe Stock.
mediaType	Media category from Adobe type IDs and format flags; mediaTypeId retains the source ID.
creatorId	Public contributor identifier when supplied.
creatorDisplayName	Public creator attribution shown with the asset.
contributorUrl	Link to the public contributor portfolio.
keywords	Asset-specific keyword strings from the public metadata.
category	Adobe Stock category name when available.
width, height	Original asset dimensions, not thumbnail dimensions.
uploadedAt	Upload timestamp if exposed; otherwise null.
isGenerativeAI	Supplied generative-AI flag; absent flags remain null.
downloadCount, viewCount	Numeric public counters if supplied; otherwise null.
licensePrice, currency	Numeric price for the default license when explicitly supplied.
previewUrl	Public preview or thumbnail link, not a licensed original.
url	The public Adobe Stock asset page.
scrapedAt	Time this metadata was observed, in UTC.
previousSeenAt	Earlier observation timestamp, added only when one exists.
downloadDelta, viewDelta	Added only when both observations contain the corresponding numeric counter.

Additional fields keep the source's pricing distinctions: licensePriceText, licenseOptions and extendedLicensePrice. A blank standard-license amount stays null even when the extended license has a visible price. An extended-license price must not be mistaken for the standard-license price. isFireflyGenerated preserves the separate source flag when present. source and sourceType identify the input that first delivered an asset.

The ASSETS and NOTES views focus on their respective columns. Their linked clean exports omit rows without those fields. Each run also saves a SUMMARY key-value record containing the delivered count, contributor count, source outcomes, duplicate exclusions, free failures and spending-limit status. Inspect SUMMARY before treating an empty dataset as evidence that a portfolio or search is empty.

Start in three steps

1. Enter topics in searchTerms, contributor portfolio links in contributorUrls, or Adobe Stock links in startUrls.
2. Choose your media filters, public sort order and global asset limit.
3. Run the Actor, inspect SUMMARY and export the ASSETS view.

The prefill searches coffee with a 20-asset limit. A cloud verification on 5 November 2026 delivered 20 distinct assets with 20 asset charges in 52 seconds. Asset-specific keywords were present, and duplicate IDs were absent. Network conditions and source responses vary. Cloud checks also completed two runs of 1,100 unique assets each, with one asset charge per row and zero source failures.

Inputs

Input	Behavior and example
searchTerms	An array such as ["coffee", "renewable energy"].
contributorUrls	Public URLs such as https://stock.adobe.com/contributor/203434592/chayathon2000.
startUrls	Public asset, keyword-search or contributor URLs on stock.adobe.com.
sortBy	relevance, newest or downloads; default relevance.
mediaTypes	Optional array: photo, illustration, vector, video, template, 3d or audio. Empty uses Adobe's combined search; select audio explicitly for music.
includeGenerativeAI	True by default. False requests Adobe's exclusion filter and excludes assets explicitly flagged as AI-generated.
compareWithPreviousRun	False by default. True remembers a completed observation for matching source/filter settings.
maxItems	Global unique-asset limit from 1 to 10,000; default 20.
proxyConfiguration	Optional proxy settings. The automatic route starts with the default datacenter pool and uses US residential only after fresh-session retries.

At least one source selector is required. Invalid values produce a free note and an invalid-input summary. They do not produce asset charges. Enter complete HTTPS links. Links to unrelated domains, account pages, download endpoints and private dashboards are not accepted as source selectors.

Sources share one global cap and are visited one page at a time in input order. A page can contribute up to 100 assets, so a smaller cap may be filled by the first source. For fixed per-query allowances, use separate runs. Repeated IDs are free.

A multi-topic scan:

{
  "searchTerms": ["coffee", "renewable energy", "healthy food"],
  "sortBy": "relevance",
  "mediaTypes": ["photo"],
  "includeGenerativeAI": false,
  "maxItems": 300
}

A repeatable contributor observation:

{
  "contributorUrls": [
    "https://stock.adobe.com/contributor/203434592/chayathon2000"
  ],
  "sortBy": "newest",
  "compareWithPreviousRun": true,
  "maxItems": 100
}

For known assets, supply their public pages in startUrls. Search sorting does not change the identity of a requested exact asset. An unavailable asset is explained without an asset charge. Keyword-search links can retain their public URL filters; the explicit sort and media settings on the Actor form take precedence where they address the same filter.

Example asset

The following values were checked against the public asset data on 5 November 2026. This example is shortened to keep the fields readable:

{
  "assetId": "94143804",
  "title": "cup of coffee latte in coffee shop",
  "mediaType": "photo",
  "creatorId": "203434592",
  "creatorDisplayName": "chayathon2000",
  "contributorUrl": "https://stock.adobe.com/contributor/203434592/chayathon2000",
  "keywords": ["hot drink", "latte", "art", "cafes", "drink", "hot", "cup", "espresso", "shop"],
  "category": "Coffee",
  "width": 4928,
  "height": 3280,
  "uploadedAt": null,
  "isGenerativeAI": false,
  "downloadCount": null,
  "viewCount": null,
  "licensePrice": null,
  "extendedLicensePrice": 79.99,
  "currency": "USD",
  "url": "https://stock.adobe.com/images/cup-of-coffee-latte-in-coffee-shop/94143804"
}

Monitor a portfolio or search again

Save your preferred input as an Apify task and attach an Apify schedule. A schedule runs the same source selectors and filters again and stores a separate dataset for each run. Webhooks can pass the successful run's dataset ID to your database, spreadsheet or automation platform. Keep the task input stable when comparing observations.

With compareWithPreviousRun: true, the first completed run establishes a baseline. Later matching runs add previousSeenAt for assets in that baseline. observationStatus distinguishes the baseline, previously observed assets and newly observed assets. changedFields lists observed changes in titles, keywords, dimensions, category, creator attribution, AI flags and license amounts. SUMMARY identifies the baseline run and counts newly observed and changed assets. The comparison is scoped to the calling Apify user, this Actor and the selected sources and filters. Completed observations replace the baseline; partial or invalid runs retain it.

An asset without previousSeenAt has no matching earlier observation in that baseline. This does not establish when it was uploaded to Adobe Stock. An asset absent from a capped or reordered result set has not necessarily been deleted. For stronger longitudinal comparisons, monitor exact asset links, retain your own historical datasets and record the cap and filters alongside them.

Counters are never reconstructed from search position or download sorting. A view or download delta is emitted only when Adobe supplies numeric values in both observations. Null is a missing value, not zero. Keep those distinctions when building charts, aggregations or alerts.

Pricing

One delivered unique asset is one Asset charge. Asset retries, duplicates, free notes and SUMMARY are not asset charges. The minimum start charge is disclosed separately. Your Apify subscription tier determines the asset rate:

Plan	Per asset	100 assets	1,000 assets
Free	$0.0005	$0.05	$0.50
Bronze	$0.000475	$0.0475	$0.475
Silver	$0.00045	$0.045	$0.45
Gold and higher	$0.0004	$0.04	$0.40

The table shows asset charges only. Add $0.0023 per GB of allocated memory, minimum one unit, on every plan. At the default 1 GB this adds $0.0023; selecting 2 GB adds $0.0046. Apify charges this start fee automatically, including runs that return only notes, empty results or failures. Those results have no asset charge. The minimum permitted run budget is $0.0028, covering the default start plus one base-rate asset; allocate a higher budget when selecting 2 GB. This start fee is higher than the earlier private-test fee of $0.00005. Asset prices and tier discounts are unchanged, and small-job totals are higher than in those earlier tests.

At the default 1 GB, total examples including the start fee are:

Plan	1 asset	2 assets	Default 20 assets	1,000 assets
Free	$0.0028	$0.0033	$0.0123	$0.5023
Bronze	$0.002775	$0.00325	$0.0118	$0.4773
Silver	$0.00275	$0.0032	$0.0113	$0.4523
Gold and higher	$0.0027	$0.0031	$0.0103	$0.4023

API and integrations

Use your Apify token in a secret environment variable; no Adobe credentials are needed.

from apify_client import ApifyClient

client = ApifyClient(token="YOUR_APIFY_TOKEN")
run = client.actor("vant47/adobe-stock-scraper").call(
    run_input={
        "searchTerms": ["renewable energy"],
        "maxItems": 100,
        "compareWithPreviousRun": True,
    }
)
for row in client.dataset(run["defaultDatasetId"]).iterate_items():
    if row.get("recordType") == "ASSET":
        print(row["assetId"], row["title"], row["keywords"])

Use each run's returned dataset ID. Export JSON or CSV, connect Make, n8n, Zapier or Sheets, and retain SUMMARY with the export for coverage context.

Use with AI agents (MCP)

An AI agent connected to https://mcp.apify.com can read the Actor's schema, choose a global cap and start a run. Name the source and wanted fields: “Collect 100 Adobe Stock coffee photos with asset-specific keywords, creator attribution and dimensions, excluding generative AI.” For portfolio work, give the public contributor link and ask the agent to inspect SUMMARY after the run.

Ask the agent to preserve nulls in any resulting table. It should not describe missing counters as zero downloads, turn newest ordering into an upload timestamp, or label a returned preview as a licensed image. The Actor's structured fields and source URLs let your agent make those distinctions explicitly.

Limits

Only public signed-out metadata is collected. The tested website routes do not expose numeric download/view counts or upload dates consistently; these fields remain null when absent. Most standard-image license prices are blank, while an extended-license amount may be visible. A listed amount is not a personalized checkout quote. Search ordering is supplied by Adobe and can change between runs. A capped portfolio observation does not prove additions or removals across the whole portfolio. Adobe can challenge or refuse anonymous traffic; exhausted retries produce free notes and partial coverage in SUMMARY. No login, account cookies, CAPTCHA solving, paid unblocking, licensed originals, contact details or private contributor dashboards are used. Public preview links remain subject to Adobe's rights and availability. Metadata is requested in English; a regional URL prefix does not select a locale. Public prices depend on the visitor context. The Actor is independently maintained and is not affiliated with Adobe.

FAQ

Are the keywords really specific to each asset? Yes. The Actor requests the asset's own public keyword list. It does not substitute suggested searches or words inferred from the title.

Can I monitor a contributor without searching a topic? Yes. Supply the public portfolio link in contributorUrls. The resulting rows retain their public creator IDs and attribution.

What does downloads ordering mean if downloadCount is null? Adobe can provide an ordering without publishing a numeric count in the same response. The Actor preserves that distinction and does not estimate a number.

Does one asset in two searches cost twice? No. Within a run, the asset ID is the deduplication key. The same asset observed in a later run is a new observation and is charged when delivered again.

Will I receive a licensed original? No. You receive metadata and public preview links. Licensing an asset remains a separate action with Adobe Stock.

Why was my result empty? Inspect SUMMARY and the NOTES view. They distinguish invalid input, empty public results, refused requests, unavailable assets and the spending limit.