Twitter (X) Media Downloader

Collect public Twitter / X media without an account, cookies or a browser. Paste post links or ask for a profile's media posts and get the actual files back — videos in any available quality, photos, GIFs, audio-only tracks and video posters — each with a public download link.

What it does

• Post mode — one link or ID per post, including x.com, twitter.com, mobile.twitter.com, /i/web/status/ links, /status/<id>/video/1 links and bare numeric IDs.
• Profile mode — walks a profile's media posts page by page, not just the handful shown to logged-out visitors.
• Both modes in one run — results are de-duplicated across the two paths.
• Status rows — deleted, protected, media-less and rate-limited targets produce an honest status row instead of a silent empty run.

Services

Service	What you get
Video download	Best, 2160p, 1080p, 720p, 480p or smallest MP4; every quality can be saved as separate files
Photo download	Original, large, medium or small image files
GIF download	The MP4 file X serves for animated GIFs
Audio extraction	M4A (AAC) or MP3 soundtrack of a video
Poster images	The video/GIF thumbnail as a separate photo file
Quoted-post media	Media inside a quoted post, marked mediaSource: "quoted post"
Author and post context	Text, date, engagement counts, author name/avatar/verification/followers, hashtags, mentions, links, reply/quote flags and quoted post IDs
Profile context	Profile ID, follower count and media count on rows found through a username
Profile filters	Minimum likes, minimum reposts and verified-authors-only

How it works

The actor reads X's public post data and the FxTwitter mirror, combines the fields each source is best at, and falls back to the X status page and oEmbed when a post looks missing. Videos and GIFs use X's direct CDN; if a direct file cannot be fetched, yt-dlp is tried as a last resort. Files are streamed to disk, checked for a sane size, stored in the run's key-value store and linked with a signed public URL. No request uses your cookies, login or API key.

Input

Field	Type	Default	Description
postUrls	array	–	Post links or bare post IDs (the prefill uses one NASA post)
usernames	array	–	Handles, @handles or profile links to scan
maxPostsPerUser	integer	20	Media posts to check per profile (1–1000)
maxItems	integer	100	Hard cap on media rows for the whole run (1–10000)
mediaTypes	array	["video", "photo", "gif"]	video, photo, gif
imageQuality	string	orig	orig, large, medium, small
videoQuality	string	best	best, 2160p, 1080p, 720p, 480p, worst
downloadAllQualities	boolean	false	One row and file per available MP4 quality
includePosterImage	boolean	true	Save each video/GIF poster as a photo
includeAltText	boolean	true	Attach X's image description when present
audioOnly	boolean	false	Save only the audio track of videos; photos and GIFs are skipped
audioFormat	string	m4a	m4a or mp3 when audio-only is on
includeQuotedMedia	boolean	true	Include media from a quoted post
includeMetadata	boolean	true	Attach post, author and engagement context
minLikes	integer	0	Profile scans only: skip posts below this like count
minRetweets	integer	0	Profile scans only: skip posts below this repost count
onlyVerifiedAuthors	boolean	false	Profile scans only: skip unverified authors
saveToFile	boolean	true	Store files and return signed links; off means metadata and CDN links only

At least one post link/ID or username is required. Directly linked posts are always collected; the three filters apply while scanning usernames.

Example: download two posts

{
  "postUrls": [
    "https://x.com/NASA/status/2107095768821576019",
    "https://x.com/SheCodeAfrica/status/2107516524965794217"
  ],
  "videoQuality": "1080p",
  "maxItems": 10
}

Example: scan a profile's media

{
  "usernames": ["NASA"],
  "maxPostsPerUser": 50,
  "maxItems": 100,
  "mediaTypes": ["video", "photo"],
  "minLikes": 100,
  "onlyVerifiedAuthors": true
}

Example: audio-only MP3

{
  "postUrls": ["https://x.com/NASA/status/2107095768821576019"],
  "audioOnly": true,
  "audioFormat": "mp3"
}

Output

Every dataset row has type: media file for a collectible file, or status for a run message.

Group	Fields
Identity	tweetId, tweetUrl, postedAt, conversationId, inReplyToTweetId, isReply, isQuote, isRetweet, quotedTweetId, quotedTweetUrl
Media	mediaType, mediaUrl, mediaUrlOriginal, contentType, mediaId, mediaKey, thumbnailUrl, posterImageUrl, filename, fileSizeBytes, width, height, durationSeconds, focusRects, allowDownload
Download	downloadUrl, downloadStatus, message
Video quality	variantQuality, variantBitrate, availableQualities, qualityCount, isBestQuality
Audio	audioFormat
Author	uploaderUsername, uploaderName, uploaderVerified, uploaderAvatarUrl, uploaderFollowers
Post context	tweetText, likeCount, retweetCount, replyCount, quoteCount, bookmarkCount, viewCount, engagementRate, hashtags, mentions, urls
Source	mediaSource (post, quoted post or poster image)
Profile	profileId, profileFollowers, profileMediaCount
Run	collectedAt

Empty fields are omitted. Status rows carry status (noInput, invalidLink, invalidUsername, notFound, rateLimited, error, noMatchingMedia, noRecords), a human-readable message, and the related tweetUrl, tweetId or username.

Example record

{
  "type": "media file",
  "tweetId": "2107095768821576019",
  "tweetUrl": "https://x.com/i/status/2107095768821576019",
  "postedAt": "2026-10-05T13:09:13+00:00",
  "mediaType": "video",
  "mediaUrl": "https://video.twimg.com/amplify_video/.../avc1/1920x1080/video.mp4?tag=16",
  "contentType": "video/mp4",
  "mediaId": "2107095744146460672",
  "filename": "2107095768821576019_video_1_1080p.mp4",
  "fileSizeBytes": 3178943,
  "width": 1920,
  "height": 1080,
  "durationSeconds": 13.3,
  "downloadUrl": "https://api.apify.com/v2/key-value-stores/{STORE_ID}/records/...?signature=...",
  "downloadStatus": "finished",
  "uploaderUsername": "NASA",
  "uploaderName": "NASA",
  "uploaderVerified": true,
  "uploaderFollowers": 92310189,
  "likeCount": 7476,
  "retweetCount": 1125,
  "quoteCount": 97,
  "bookmarkCount": 332,
  "viewCount": 2961506,
  "variantQuality": "1080p",
  "isBestQuality": true,
  "mediaSource": "post",
  "collectedAt": "2026-10-29T00:00:00+00:00"
}

Usage cost

You pay the actor's per-result price shown on its Apify page plus normal Apify platform usage. With saveToFile: false no media bytes are stored, so a metadata-only run is the cheapest option. Downloads add key-value store usage proportional to the real file sizes. Keep maxItems and maxPostsPerUser at the values you need — the actor stops as soon as maxItems is reached.

FAQ

Do I need an X account, cookies or an API key? No. Everything is read from public, logged-out sources.

Can I download from a private or protected account? No. Protected and deleted posts return a notFound status row; they are not silently skipped.

How far back does a profile scan go? The walk follows X's media timeline page by page until maxPostsPerUser, maxItems or the end of the timeline. Large profiles can be rate-limited; lower maxPostsPerUser and re-run later.

How are video qualities chosen? The closest available quality to your preference is used. Portrait videos are compared by their short side, so a 1080×1920 video counts as 1080p. With downloadAllQualities, every progressive MP4 becomes its own row and file.

What happens with quoted posts? Media inside a quoted post is included by default and marked mediaSource: "quoted post". Turn includeQuotedMedia off to keep only the linked post's own media.

Does audio-only work on GIFs? No. Animated GIFs have no audio track, so they are skipped in audio-only mode along with photos.

What if a post has no media? You get one noMatchingMedia status row explaining it, not an empty dataset.

Do download links expire? Signed key-value store links follow your Apify storage retention. Re-run the actor or copy the files to your own storage for permanent archiving.

How do the profile filters treat unknown counts? A post whose like/repost count or verification status is unknown is kept rather than excluded.

Can I still use my saved inputs from the previous version? Yes. Old input keys are mapped automatically in code, so saved tasks keep running.

Notes

• All timestamps are UTC ISO 8601.
• Engagement rate is derived from likes, reposts, replies and quotes divided by the author's follower count; it is omitted when the follower count is unknown.
• Media URLs are direct X CDN links and open without a Referer header.
• Download links are public signed links tied to the run's key-value store.
• Use the actor only for content you have the right to download.

Limitations

• Public posts only: deleted posts, protected accounts and login-gated content are not accessible.
• Profile scans cover the most recent media posts X exposes publicly, up to 1,000 posts per run; deeper history is not addressable and very large profiles can hit rate limits, so re-run later if a scan stops early.
• A single media file larger than 1 GB is rejected to keep memory bounded.
• Some media is marked by X as not downloadable; those rows report downloadStatus: "notDownloaded" with an explanation.
• Quote media depends on the quoted post still being public.
• X changes its public endpoints periodically; the actor keeps several fallbacks, but a run can still return rateLimited or error status rows during an outage.