Universal Media Downloader — video & audio from 1800+ sites

Download video or audio from almost any URL over a simple HTTP API. Powered by yt-dlp + ffmpeg, it supports 1800+ sites — YouTube, X/Twitter, TikTok, Instagram, Vimeo, SoundCloud, Reddit, Twitch, Facebook, and many more — plus direct .mp4/.mp3 links. Choose the format and quality, extract just the audio, or read a URL's metadata without downloading. No binaries to install, no server to run.

• ⬇️ Any site → MP4 / WebM / MKV, or audio only → MP3 / M4A / WAV / Opus / …
• 🎚️ Pick quality (480p → 4K) and a max file size
• 🔎 Info mode: title, duration, thumbnail and available formats — no download, cheaper
• 🌐 Optional residential proxy for sites that block datacenter IPs (e.g. YouTube)
• ⚡ HTTP API (Standby) + /async for big files + usable from AI agents (MCP)
• 💵 Pay per download + per 100 MB (residential proxy billed separately)

Sample output

{
  "ok": true,
  "url": "https://api.apify.com/v2/key-value-stores/…/records/media_lq3k.mp4",
  "contentType": "video/mp4",
  "sizeMB": 4.8,
  "billed100MB": 1,
  "kind": "video",
  "ext": "mp4",
  "proxied": false
}

How to use

HTTP API (Standby — recommended)

Base URL: https://synthetic-ia--media-downloader.apify.actor with header Authorization: Bearer <APIFY_TOKEN>.

Endpoint	Body	Does
POST /download	{ url, format?, audioOnly?, maxHeight?, maxFileSizeMb?, proxy? }	Download → file URL
POST /info	{ url, proxy? }	Metadata + available formats (no download)
POST /async	{ ...params } → { runId }	Start a big download; then GET /status/<runId>
GET /	—	Endpoint list as JSON

# Best-quality MP4 up to 1080p
curl -X POST ".../download" -H "Authorization: Bearer $APIFY_TOKEN" \
  -d '{"url":"https://vimeo.com/…","maxHeight":1080}'

# Audio only → MP3
curl -X POST ".../download" -H "Authorization: Bearer $APIFY_TOKEN" \
  -d '{"url":"https://soundcloud.com/…","audioOnly":true,"format":"mp3"}'

# A site that blocks servers (YouTube) → use the residential proxy
curl -X POST ".../download" -H "Authorization: Bearer $APIFY_TOKEN" \
  -d '{"url":"https://www.youtube.com/watch?v=…","proxy":"residential"}'

One-shot run

Set url (and options) and run once. The file is saved to the run's key-value store as OUTPUT and a record with the URL is pushed to the dataset.

Input

• url — the page or file to download from.
• operation — download or info (metadata only).
• audioOnly — extract just the audio.
• format — video mp4|webm|mkv, audio mp3|m4a|wav|opus|aac|flac|ogg.
• maxHeight — 0 (best), 480, 720, 1080, 2160.
• maxFileSizeMb — abort above this (1–1024).
• proxy — none (direct, default), datacenter, or residential.

About the proxy (important for YouTube)

Requests run from Apify's datacenter IPs, which some sites — most notably YouTube — block. Set proxy: "residential" to route the download through real home IPs so the site treats you like a normal user. Residential bandwidth is billed per GB, so it's charged separately (see Pricing). For most other sites, leave proxy as none.

Pricing

What	Price (Free tier)
Per download	$0.05
Per 100 MB of output	$0.10
Residential proxy, per 100 MB	$2.00 (only when used)

Prices drop on higher usage tiers. A 5-minute 1080p video (~80 MB) direct ≈ $0.05 + $0.10 = $0.15. Info mode is charged as a normal download event only if it downloads; pure metadata is light. Datacenter/no-proxy downloads never pay the residential fee.

Related Actors

• ffmpeg API — transcode, compress, trim, GIF, thumbnail the downloaded file.
• Live Stream Clipper — clip a live stream instead of downloading a full video.
• AI Voice Generator & Cloner — turn text into speech, or clone a voice.

Limits

• Output capped at 1024 MB per download; use /async for big files (30-min timeout).
• http(s) URLs only; private/loopback hosts are blocked.

Legal

You are responsible for having the right to download the content you request. Use it for your own media, content you're licensed to use, or content that permits downloading. Do not use it to infringe copyright or violate a site's terms of service.