Pexels Stock Media Scraper

Scrape Pexels photos and videos without a Pexels account, cookie, or API key. Search by keyword, browse the curated feed, pull a photographer's portfolio or a collection, look up exact photos and videos by ID or URL, find similar media, or paste any Pexels link. Get direct CDN URLs at every size and video rendition, tags, colors, dimensions, photographer credit, engagement counts, camera metadata, and the full photographer profile.

What this actor does

• Seven ways in: search, curated, byMediaId, byUser, byCollection, similarMedia, byUrl
• Photos and videos in one actor: search one type or both, and mix photo/video IDs and URLs
• Full filter surface: orientation, size, color or hex, people count, people age, locale, minimum width/height, minimum FPS, duration range, and publish-date window
• Direct media URLs: multiple image sizes, the original file, every video rendition, and a highest-resolution convenience URL
• Photographer credit: name, profile, avatar, location, socials, donation link, hero/superhero badges, and full profile statistics
• Engagement and camera data: views, downloads, likes, image views, reactions, EXIF camera fields, content-safety ratings, and file size when includeDetails is on
• Empty fields are omitted: you never have to defend against null
• No browser: the actor uses Pexels' public web JSON route with an HTML fallback, so runs are small and fast

Modes

Mode	What you give it	What you get
search	searchQueries + mediaType	Keyword results, photos, videos, or both
curated	nothing	Pexels' editorial photo feed (paginated)
byMediaId	mediaIds	Exact photos and videos by ID or URL, enriched with detail data
byUser	username	A photographer's full portfolio, with profile stats
byCollection	collectionSlug	Every photo and video in a collection
similarMedia	mediaIds	Pexels' related-media feed for each seed
byUrl	startUrls	Automatically routes search, photo, video, profile, collection, image, and video-file URLs

Legacy mode values are still accepted in code: byPhotoId maps to byMediaId and similarPhotos maps to similarMedia. The old searchQuery and photoIds input keys are mapped too, so saved tasks keep working.

Input

Field	Type	Default	Description
mode	string	search	search / curated / byMediaId / byUser / byCollection / similarMedia / byUrl
searchQueries	array	[]	One or more keywords for search
mediaType	string	photos	photos / videos / all
startUrls	array	[]	Pexels URLs for byUrl
orientation	string	all	all / landscape / portrait / square
size	string	all	all / large (24 MP or 4K) / medium (12 MP or Full HD) / small (4 MP or HD)
color	string	all	12 Pexels colors, photos only
colorHex	string	empty	Any hex color such as #2563eb; overrides color
peopleCount	string	all	all / 0 / 1 / 2
peopleAge	string	all	all / baby / child / teenager / adult
locale	string	en-US	28 Pexels locales; localizes titles, descriptions, and search
minWidth	integer	0	Minimum width in pixels (0 = off)
minHeight	integer	0	Minimum height in pixels (0 = off)
minFps	integer	0	Minimum video frame rate (0 = off); photos always pass
minDuration	integer	0	Minimum video duration in seconds (0 = off)
maxDuration	integer	0	Maximum video duration in seconds (0 = off)
publishedAfter	string	empty	YYYY-MM-DD; keep media published on or after this date
publishedBefore	string	empty	YYYY-MM-DD; keep media published on or before this date
mediaIds	array	[]	Photo/video IDs or URLs for ID and similar modes
username	string	empty	Photographer handle, ID, or profile URL
collectionSlug	string	empty	Collection slug or full URL
maxItems	integer	30	Global cap (1-5000); prefill 3 for the daily test
includeDetails	boolean	false	Add views/downloads/likes, EXIF, content safety, and full profile stats; one extra request per item
dedupe	boolean	true	Keep only the first occurrence of each media ID
proxyConfiguration	object	off	Optional proxy override; the actor works directly and escalates only when blocked

Examples

Search landscape nature photos at least 24 MP:

{
  "mode": "search",
  "searchQueries": ["nature"],
  "mediaType": "photos",
  "orientation": "landscape",
  "size": "large",
  "maxItems": 50
}

Search short ocean videos at 60 fps or higher:

{
  "mode": "search",
  "searchQueries": ["ocean"],
  "mediaType": "videos",
  "minFps": 60,
  "maxDuration": 30,
  "maxItems": 20
}

Get a photographer's portfolio in Spanish:

{
  "mode": "byUser",
  "username": "alex-ivanov-773258579",
  "locale": "es-ES",
  "maxItems": 100
}

Look up a photo and a video by ID:

{
  "mode": "byMediaId",
  "mediaIds": ["18872916", "25961000"]
}

Scrape a collection, including its videos:

{
  "mode": "byCollection",
  "collectionSlug": "fashion-kwvfk6o",
  "mediaType": "all",
  "maxItems": 100
}

Find media similar to a photo or video:

{
  "mode": "similarMedia",
  "mediaIds": ["18872916"],
  "maxItems": 24
}

Paste any Pexels link:

{
  "mode": "byUrl",
  "startUrls": [
    {"url": "https://www.pexels.com/search/nature/?orientation=landscape"},
    {"url": "https://www.pexels.com/photo/18872916/"},
    {"url": "https://www.pexels.com/@alex-ivanov-773258579/"},
    {"url": "https://www.pexels.com/collections/aquatic-aesthetic-bw7dkse/"}
  ],
  "maxItems": 60
}

Output

Every row is one photo or video. Existing field names are preserved.

Photo fields

photoId, title, description, alt, slug, width, height, orientation, aspectRatio, license, licenseUrl, featured, createdAt, publishedAt, updatedAt, feedAt, photographerName, photographerId, photographerUsername, photographerUrl, photographerAvatarUrl, photographerLocation, photographerInstagram, photographerTwitter, photographerIsHero, photographerSuperhero, photographerActiveHero, photographerBadges, photographerDonateUrl, attribution, tags, imageSmallUrl, imageMediumUrl, imageLargeUrl, imageOriginalUrl, downloadUrl, downloadPageUrl, avgColor, colorPalette, collectionIds, reactionEmotionalCount, reactionTechnicalCount, reactionStylisticCount, pageUrl, sourceUrl, recordType, scrapedAt.

With includeDetails or byMediaId, photos can also include cameraMake, cameraModel, cameraIso, cameraAperture, cameraFocalLength, cameraShutterSpeed, cameraSoftware, photographedAt, cameraOrientation, fileSizeBytes, latitude, longitude, locationName, copyright, googlePlaceId, contentSafety, viewsCount, downloadsCount, likesCount, imageViewsCount, and the full profile fields below.

Video fields

videoId, title, description, slug, width, height, orientation, aspectRatio, durationSeconds, fps, license, licenseUrl, featured, createdAt, publishedAt, updatedAt, feedAt, photographerName, photographerId, photographerUsername, photographerUrl, photographerAvatarUrl, photographerLocation, photographerInstagram, photographerTwitter, photographerIsHero, photographerSuperhero, photographerActiveHero, photographerBadges, photographerDonateUrl, attribution, tags, videoPreviewUrl, videoProvider, thumbnailSmallUrl, thumbnailMediumUrl, thumbnailLargeUrl, downloadUrl, downloadPageUrl, videoFiles[], bestVideoUrl, collectionIds, reactionEmotionalCount, reactionTechnicalCount, reactionStylisticCount, pageUrl, sourceUrl, recordType, scrapedAt.

videoFiles[] entries contain url, quality (when Pexels supplies it), width, height, fps, fileType, and downloadUrl. There is no top-level videoQuality field; the quality label lives inside each videoFiles[] entry.

Profile fields (detail / portfolio modes)

photographerBio, photographerWebsite, photographerYoutube, photographerTiktok, photographerFollowersCount, photographerFollowingsCount, photographerPhotosCount, photographerMediaCount, photographerPortfolioMediaCount, photographerCollectionsCount, photographerFirstUploadedOn, photographerLastUploadedAt, photographerImageViewsAllTime, photographerImageViewsLast30Days, photographerChallengesWonCount, photographerChallengesParticipatedCount, photographerHomepageFeaturedCount, photographerLeaderboardLast30DaysPosition, photographerLeaderboardAllTimePosition, photographerConfirmed, photographerBadges[].

Provenance

foundBy is one of keyword, curated, id, user, collection, or similar; query, seedMediaId, collectionSlug, and username are added where they apply. sourceUrl remains the media page URL for backward compatibility.

Limits and data notes

• Search depth: Pexels serves at most about 480 records per search query (20 pages of 24). Page 21 and beyond return empty even though the site displays total_results=10000. Use narrower queries, filters, multiple queries, or the curated/profile/collection modes for more.
• Curated feed: photos only. The public feed contains about 68,000 curated photos and paginates at 80 per page.
• Photographer, collection, and similar feeds: 80 records per page and paginate to the source total. The HTML fallback is limited to its first 12 records because ?page=N is ignored on those pages; the actor uses the public web JSON route for full pagination.
• Video by ID: Pexels returns 403 for a bare numeric video URL without a locale prefix; the actor resolves video IDs through the media JSON route first, then falls back to the locale-prefixed video page (/{locale}/video/<id>/) and its detail data.
• Server-side filters: orientation, size, color/hex, people count, people age, locale, and video resolution/frame-rate parameters are honored. Duration, width, height, and publish-date filters are applied by the actor on the returned records because Pexels' web route ignores them.
• Camera and location fields are conditional: in a 10-photo sample, camera data was present on 50%, a location name on 20%, and latitude/longitude on 0%. A later 50-photo audit found camera and location text on 26-60% of records but no coordinates. Empty fields are omitted, never fabricated.
• Coordinates and collection membership are not public: the anonymous Pexels route exposes mediumDetails latitude/longitude as null and returns an empty collection_ids array. The actor keeps latitude, longitude, and collectionIds for compatibility but omits them when Pexels does not provide values.
• Image originals: most originals are JPEG, but some are PNG. The actor takes the original URL from Pexels' own download link instead of assuming an extension.
• Attribution: Pexels asks API users to credit photographers and link back to Pexels. The actor adds a ready-to-use attribution field and licenseUrl to every record. The Pexels license itself does not require attribution.
• No account, cookie, or API key: all data comes from public Pexels pages and Pexels' public browser route. The actor keeps an HTML fallback if the public route changes.
• Proxy: direct connections are the default. The actor escalates automatically only if Pexels blocks the run. The optional proxyConfiguration field is an override, not a requirement.

Use cases

• Content and social teams sourcing campaign imagery and clips
• Designers building moodboards and placeholders at specific sizes and orientations
• Video editors finding 4K/60 fps clips by duration
• AI and ML teams building tagged image/video datasets with dimensions, colors, and engagement signals
• Digital asset teams mirroring direct CDN URLs into a DAM with photographer credit and license traceability
• Localization teams pulling the same query in multiple languages

FAQ

Is this affiliated with Pexels? No. This is an independent third-party actor that reads public Pexels pages. It is not affiliated with or endorsed by Pexels.

Do I need a Pexels account or API key? No. The actor does not use the official Pexels API, so there is no account, API key, quota, or cookie to supply.

Are the media files free to use? Pexels publishes its content under the Pexels License, which allows free personal and commercial use without attribution. Review the current license and the context of each asset before publishing. The actor returns licenseUrl and attribution to make credit easy.

Why do some records miss camera or location fields? Those values come from the original upload and the photographer's profile. Pexels only retains them when they were supplied. The actor omits unavailable fields instead of returning an empty column.

Why did search stop at about 480 records? That is Pexels' anonymous search ceiling. Split the topic into narrower queries or use curated, byUser, byCollection, or similarMedia, which paginate to larger totals.

Why does a collection return 84 records when I asked for 100? The collection itself has 84 records. maxItems is a cap, not a guarantee that the source contains that many matches.

Can I get videos in a collection or portfolio? Yes. Set mediaType to videos or all. Collection media type filtering is applied by the actor; portfolio media type is also filtered server-side.

What is includeDetails? It opens each media's detail page to add views, downloads, likes, camera/EXIF data, content safety, and full profile stats. It costs one extra request per item. byMediaId always enriches because exact lookup is expected to return the complete record.

Can I paste a normal Pexels URL? Yes. byUrl accepts search pages (with their filters), photo pages, video pages, photographer profiles, collections, image-file URLs, and video-file URLs.

Are the CDN links stable? All tested images.pexels.com and videos.pexels.com links returned 200. The actor returns the URLs exactly as Pexels supplies them, including the original file and every video rendition. Keep the media ID as the stable key.

How fresh is the data? Every run reads Pexels live, so engagement counts and search order reflect the moment of the run.