Download Files from URLs - Images, PDFs, Videos & ZIP

Give it a list of file URLs, or a dataset from another Actor plus the name of the field that holds the URLs. It downloads every file and saves it to an Apify key-value store. Each file gets one row with a direct link to the saved copy, the file name, type, size in bytes and SHA-256. You can also ask for everything in a ZIP.

• Any file type. Images, PDFs, videos, audio, Office documents, archives. The type is read from the file's first bytes, not only from the URL, so a link without an extension still gets the right name and type.
• Straight from a scraper's results. Point it at a dataset (for example the output of a product or listing scraper) and the field with the image or file URL. Fields that hold a list of URLs, or objects with url, src or href, are read in full. Nested fields work with dots (media.0.url).
• One file or one ZIP. files saves each file as its own record with its own link. zip packs everything into files.zip without compression, so it is fast and every file stays byte-for-byte the same. both does both. A large ZIP is split into parts (files-part2.zip, ...) so it always fits the run's disk.
• Large files are streamed. Files are written to disk and uploaded from there, not held in memory. Measured on Apify at 256 MB memory: a 113 MB TIFF and a 134 MB TIFF in one run, saved as files and as a 2-part ZIP.
• Check the copy. Every row has the SHA-256 and the size of the saved file, so you can check it against the source or spot duplicate files.
• Clear file names. The file name comes from the URL or the server's Content-Disposition. Two different files with the same name get -2, -3, ... Unsafe characters are replaced.
• Not charged for files it could not save. 404, a web page instead of a file, too large, wrong type, blocked by robots.txt, needs a sign-in, no connection: each comes back as a free row that says why. A run where no file could be saved is not charged at all, not even the start fee.
• Only the URLs you give it. It does not crawl sites or follow links inside pages.

Measured on Apify at 256 MB memory: 100 Wikimedia Commons photos (23 MB) in 43 s; 3 files of 250 MB in total with both in 90 s.

Input

Field	Type	Default	What it does
urls	array of URLs	(example photo)	Direct links to the files, one per line. If this and the dataset are both left out, one example photo from Wikimedia Commons is downloaded and the log says so.
datasetId	dataset	none	An Apify dataset whose rows hold file URLs, for example a scraper's results. Its URLs are added to urls.
datasetField	text	url	The field of each dataset row that holds the URL (imageUrl, images, media.0.url, ...).
saveAs	files / zip / both	files	Separate records, one ZIP (split into parts when large), or both.
fileTypes	array	all types	Keep only these kinds: image, video, audio, pdf, document, archive, other, or extensions such as jpg, mp4, docx. Other files come back as free type-not-allowed rows.
skipWebPages	boolean	true	An address that returns an HTML page instead of a file is not saved (free web-page row). Off: the HTML is saved like any other file.
outputKeyValueStoreId	key-value store	this run's store	Where to save the files. Choose your own store to keep the files and links after your plan's data retention period.
maxFileMegabytes	1-4000	500	A file larger than this is not saved (free too-large row).
requestTimeoutSecs	5-600	60	A download that receives no data for this long is stopped and tried again. A slow but moving download is not stopped.
maxConcurrency	1-16	4	Files downloaded at the same time. Lower it if a site answers 429.

Example:

{
  "datasetId": "<your scraper's dataset id>",
  "datasetField": "imageUrl",
  "saveAs": "both",
  "fileTypes": ["image"]
}

Output

One dataset row per URL. outputUrl opens the saved file directly (a signed link, no API token needed). zipUrl and zipEntry say which ZIP part holds it.

{
  "status": "ok",
  "position": 2,
  "url": "https://upload.wikimedia.org/wikipedia/commons/f/f0/Paris_view_from_Eiffel.jpg",
  "finalUrl": "https://upload.wikimedia.org/wikipedia/commons/f/f0/Paris_view_from_Eiffel.jpg",
  "outputUrl": "https://api.apify.com/v2/key-value-stores/<store>/records/Paris_view_from_Eiffel.jpg?signature=<signature>",
  "key": "Paris_view_from_Eiffel.jpg",
  "keyValueStoreId": "<store>",
  "fileName": "Paris_view_from_Eiffel.jpg",
  "contentType": "image/jpeg",
  "fileType": "image",
  "extension": "jpg",
  "bytes": 2762548,
  "sha256": "4b4a6566a60033ac051e2b5a6cc8a228e165630738e73fe046b3f43c5e6e83e8",
  "zipUrl": "https://api.apify.com/v2/key-value-stores/<store>/records/files.zip?signature=<signature>",
  "zipEntry": "files.zip/Paris_view_from_Eiffel.jpg",
  "httpStatus": 200,
  "redirects": 0,
  "note": null
}

All columns: status, position (order in your input), url, finalUrl (after redirects), outputUrl, key, keyValueStoreId, fileName, contentType, fileType, extension, bytes, sha256, zipUrl, zipEntry, httpStatus, redirects, durationMs, note, downloadedAt.

The same URL given twice is downloaded and charged once (the log says how many were skipped).

Free rows (not charged)

status	When
not-found	The site says the file does not exist (404 or 410).
web-page	The address returned an HTML page, not a file (with skipWebPages on).
type-not-allowed	The file is not one of the fileTypes you chose.
too-large	Larger than maxFileMegabytes, or larger than the run's temporary disk (the row says to run with more memory).
empty-file	The server sent a file with no content.
robots-disallowed / robots-unreachable	The site's robots.txt does not allow automated reading of that address, or could not be read.
login-required / blocked	The file needs a sign-in, or the site refused the reader or showed a check page. It does not sign in, solve check pages or switch to proxies.
http-error / unreachable / unreadable / download-failed	Another HTTP error, no connection, no usable answer after retries (429 and 5xx are retried), or a download that broke off.
storage-failed	The file was downloaded but could not be saved to the key-value store.
bad-input	Not an http(s) URL, a private or internal network address, or a dataset field that holds no URL (the row says why).
time-limit / budget-reached	The run was close to its timeout or to the maximum total charge you set; the remaining files were not started and not charged.

Pricing

Pay per event:

• File downloaded: $0.0049 per file on the Free plan, down to $0.0035 on higher plans. Charged only when the file is saved and its row is returned. Includes the first 10 MB of the file.
• Each further 10 MB: $0.00295 per 10 MB above the first 10 MB on the Free plan, down to $0.0023. A 25 MB video counts 2 extra blocks; a 4 MB photo counts none. This is counted per stored copy: with Save as = both (separate files and the ZIP) each file is stored twice, so a 25 MB video counts 4 extra blocks. Files under 10 MB cost the same in every mode.
• Run start: $0.006 per run on the Free plan, down to $0.0044. Charged once, only in a run that saves at least one file.

Platform usage (compute, storage writes and data transfer) is included; you do not pay it on top. Before each file is downloaded, the Actor checks that the maximum total charge you set has room for it; a file that would go over is not downloaded and comes back as a free budget-reached row.

Examples on the Free plan: 1,000 product photos under 10 MB each = $4.906. One 100 MB video = $0.006 + $0.0049 + 9 x $0.00295 = $0.038 as files or as a ZIP, or $0.006 + $0.0049 + 18 x $0.00295 = $0.064 with both.

Memory and large files

The default memory is 256 MB. Memory is not what limits file size, because files are streamed. Disk is: the run's temporary disk grows with its memory (a run at 256 MB memory had 512 MB of disk, and the log prints the free disk at start). Measured at 256 MB: files up to 134 MB each, 250 MB in one run, saved as files and ZIP. For single files larger than about 200 MB, run with more memory. If a file does not fit, it comes back as a free too-large row that says so; the run does not crash and the other files are kept.

Use it from code

curl -X POST "https://api.apify.com/v2/acts/neverempty~download-files-from-urls/run-sync-get-dataset-items?token=<YOUR_TOKEN>" \
  -H "Content-Type: application/json" \
  -d '{"urls":["https://example.com/report.pdf","https://example.com/photo.jpg"]}'

The response is the dataset rows; outputUrl is the saved file.

Notes

• Files are fetched with a plain HTTP client from Apify's own network, following each site's robots.txt. No proxies are used to get around a refusal.
• Addresses that point to private or internal networks are not requested.
• Only download files you have the right to copy.

Support

Found a problem or need another option? Open an issue on the Issues tab of this Actor.