Media Downloader - All-in-One Video, Audio and Metadata

What is Media Downloader - All-in-One Video, Audio and Metadata?

Media Downloader is an Apify Actor for creators, marketing teams, archivists, and AI-agent builders who need to save media they own or are permitted to use. Paste public media page URLs from platforms such as YouTube, Vimeo, TikTok, Instagram, Reddit, SoundCloud, Dailymotion, Pinterest, or X, and the Actor saves a playable video or audio file to Apify key-value storage. It also writes one clean dataset row per item with title, creator, duration, quality, file key, and a download link. It is usable through Apify MCP, so an agent can fetch a permitted clip and hand the file link to the next step.

Best fit and connected workflows

This Actor fits well when you want to:

• archive your own published videos or podcasts from several platforms into one storage location
• collect licensed or permitted clips for an editing, captioning, or review pipeline
• save audio-only copies for transcription or accessibility work
• read media metadata (title, creator, duration, format, thumbnail) without downloading the file, using metadataOnly
• give an AI agent a bounded, permission-confirmed way to fetch a media file by URL

Each row records the source URL, the saved file key, and a status, so downstream steps know exactly which files exist.

Practical scenario

Priya manages a training team that publishes product walkthroughs on YouTube and Vimeo. Before a platform migration, she pastes the twenty walkthrough URLs, confirms she owns the content, and runs the Actor in video mode at 720p. The dataset lists each walkthrough with its title, duration, file size, and a key-value store link. She downloads the files from the store, checks the durations against her content plan, and uploads them to the new learning platform.

Input fields

Field	Use it for
startUrls	Media page URLs to process. Required.
authorizationConfirmed	Confirms you own the media or have permission to download it. Required and must be true.
downloadMode	video, audio, or metadataOnly.
preferredQuality	Maximum video height: 360p, 480p, 720p, 1080p, or best.
preferredContainer	mp4, webm, or any for video files.
audioFormat	m4a, mp3, or best in audio mode.
includePlaylists	Expand playlist URLs into individual items, up to maxItems.
maxItems	Cap on URLs or playlist entries per run.
maxDurationMinutes	Skip media longer than this.
maxFileSizeMb	Skip files larger than this.
includeMetadataJson	Save the full extractor metadata JSON for each item.
cookiesTxt	Optional cookies for media you are allowed to access with your own session.
proxyConfiguration	Apify Proxy settings; residential proxies are the default.

Focused JSON input example

{
  "startUrls": [{ "url": "https://www.youtube.com/watch?v=abc123XYZ00" }],
  "authorizationConfirmed": true,
  "downloadMode": "video",
  "preferredQuality": "720p",
  "preferredContainer": "mp4",
  "maxItems": 5,
  "maxFileSizeMb": 200
}

Output fields

Field	Meaning
sourceUrl, parentUrl, canonicalUrl	The URL you supplied, the playlist it came from, and the canonical page.
platform, extractor, mediaId	Detected platform and media identifier.
title, creator, creatorUrl	Public title and creator.
durationSeconds, viewCount, uploadDate, thumbnailUrl	Public media metadata when the platform exposes it.
mediaType, downloadMode, preferredQuality	What was requested and saved.
formatId, formatExtension, qualityLabel, width, height	The selected media format.
directMediaUrl	An expiring direct media URL in metadata mode, when available.
fileKey, fileUrl, fileName, fileExtension, fileSizeBytes, contentType	The saved file in the key-value store.
metadataKey, metadataUrl	The saved metadata JSON record.
status, warnings	Per-URL result and notes; a diagnostic message is included when an item is skipped.
scrapedAt	Processing time.

Illustrative dataset record

{
  "sourceUrl": "https://www.youtube.com/watch?v=abc123XYZ00",
  "canonicalUrl": "https://www.youtube.com/watch?v=abc123XYZ00",
  "platform": "youtube",
  "mediaId": "abc123XYZ00",
  "title": "Product walkthrough: getting started",
  "creator": "Example Studio",
  "durationSeconds": 214,
  "uploadDate": "20250205",
  "mediaType": "video",
  "downloadMode": "video",
  "preferredQuality": "720p",
  "formatExtension": "mp4",
  "qualityLabel": "1280x720",
  "width": 1280,
  "height": 720,
  "fileKey": "video-abc123XYZ00-product-walkthrough.mp4",
  "fileName": "product-walkthrough.mp4",
  "fileSizeBytes": 18734120,
  "contentType": "video/mp4",
  "metadataKey": "metadata-abc123XYZ00.json",
  "status": "success",
  "warnings": [],
  "scrapedAt": "2026-10-01T12:56:02Z"
}

The values above are illustrative. Real rows also include fileUrl and metadataUrl links into the run's key-value store.

How it works

The Actor checks that authorizationConfirmed is true, then probes each URL with the open-source yt-dlp extractor. It applies your duration limit, picks the best format within your quality and container preferences, downloads the file, and checks it against maxFileSizeMb. Video and audio files are saved to the run's key-value store with predictable keys such as video-{mediaId}-{slug}.mp4, and the dataset row links to them. In metadataOnly mode the Actor stops after the probe and returns the metadata plus an expiring direct media URL when the platform provides one.

The key-value store also holds OUTPUT and RUN_SUMMARY records with the run outcome: COMPLETE, PARTIAL, VALID_EMPTY, INVALID_INPUT, UPSTREAM_FAILED, or CONFIG_ERROR.

Pricing

This Actor uses Pay per event plus Apify platform usage.

• apify-actor-start: charged when the run starts.
• media-downloaded: charged for each video or audio file saved to the key-value store.
• metadata-extracted: charged for each URL successfully probed in metadataOnly mode.

Only saved files and successful metadata probes are charged, and maxItems caps the number of billable items. Larger files and higher quality settings use more platform storage and transfer, so start at 720p unless you need more. Check the live Pricing tab for current event prices.

Use with AI agents (MCP)

The Actor is available through Apify MCP as khadinakbar/all-in-one-media-downloader. A typical agent request:

I own this video: https://www.youtube.com/watch?v=abc123XYZ00. Save it as an mp4 at 720p with authorization confirmed, then give me the file link, title, and duration from the dataset row and the run outcome from OUTPUT.

The agent reads fileUrl for the saved file, title and durationSeconds for context, and status for the per-item result. sourceUrl and scrapedAt provide provenance, and maxItems keeps playlist runs bounded in size and cost.

Run with the Apify API

import { ApifyClient } from 'apify-client';

const client = new ApifyClient({ token: process.env.APIFY_TOKEN });

const run = await client.actor('khadinakbar/all-in-one-media-downloader').call({
  startUrls: [{ url: 'https://www.youtube.com/watch?v=abc123XYZ00' }],
  authorizationConfirmed: true,
  downloadMode: 'audio',
  audioFormat: 'm4a',
});

const { items } = await client.dataset(run.defaultDatasetId).listItems();
for (const item of items) {
  console.log(item.status, item.title, item.fileUrl);
}

from apify_client import ApifyClient
import os

client = ApifyClient(os.environ["APIFY_TOKEN"])

run = client.actor("khadinakbar/all-in-one-media-downloader").call(
    run_input={
        "startUrls": [{"url": "https://www.youtube.com/watch?v=abc123XYZ00"}],
        "authorizationConfirmed": True,
        "downloadMode": "metadataOnly",
    }
)

for item in client.dataset(run["defaultDatasetId"]).iterate_items():
    print(item["title"], item["durationSeconds"], item["directMediaUrl"])

Best results and outcome guidance

Start with one URL in metadataOnly mode to confirm the platform is recognised, then switch to video or audio. Keep preferredQuality at 720p and set maxFileSizeMb for agent workflows that pass files onward. Use includePlaylists only when you expect several files, with a focused maxItems. Residential proxies are the default because media platforms rate-limit datacenter traffic. For media you can open only with your own login, provide your own cookiesTxt. A PARTIAL outcome means some files were saved and others were skipped; each row's status shows which.

Continue the workflow

• Then use YouTube Transcript and Subtitle Data Scraper when your next step needs the spoken text of a YouTube video rather than the media file.
• Then use YouTube Comments Scraper (with Replies) when you want audience reactions to the same videos you archived.

Design note

I designed the Actor to ask for an explicit authorization flag and to default it to false, so an accidental run saves nothing. I also found that predictable storage keys matter more than file names for agents, which is why every row carries the exact fileKey of the saved file.

FAQ

Which sites are supported?

Most public media pages that the yt-dlp extractor recognises, including YouTube, Vimeo, TikTok, Instagram, Reddit, SoundCloud, Dailymotion, Pinterest, and X.

Where do the files go?

Into the run's key-value store. Each dataset row has the fileKey and a fileUrl link to download it.

Can I get only metadata?

Yes. Set downloadMode to metadataOnly to get metadata and, when available, an expiring direct media URL.

Can I download a whole playlist?

Yes. Turn on includePlaylists and set maxItems to the number of entries you want.

Responsible use

Download only media you own, have the rights holder's permission to use, or are otherwise legally allowed to save. This Actor respects copyright, paywalls, DRM, and private-account boundaries, and it is intended for permitted public media under applicable law and each platform's site terms.