Download, transcribe, and get metadata for any video/audio URL. Managed yt-dlp + speech-to-text.
Runs on your machine from its package; the steps are below.
At most: only reads or looks up information.
Odel charges nothing per call. The publisher may charge for its own service.
Runs locally over stdio
This server isn't hosted — your MCP client launches it from a package registry. Use one of the commands below, or drop the config into your client (e.g. Claude Desktop).
Run
npx -y importly-mcp
MCP client config
{
"mcpServers": {
"importly": {
"command": "npx",
"args": [
"-y",
"importly-mcp"
]
}
}
}Give your AI agent the ability to download, transcribe, and inspect any video or audio URL — YouTube, TikTok, X, podcasts, and 1000+ other sites.
Importly runs yt-dlp, rotating residential proxies, ffmpeg, and speech-to-text server-side, so your agent (and you) never deal with IP bans, 429s, cookie rot, or extractor breakage.
| Tool | What it does |
|---|---|
import_media / check_import | Download media to hosted storage, get a direct link (optionally with a transcript) |
transcribe_media / check_transcription | Speech-to-text with timestamped segments |
get_metadata / check_metadata | Full metadata: formats, uploader, thumbnails, stats |
get_basic_metadata / check_basic_metadata | Fast title/duration/thumbnail lookup |
check_balance | Remaining credit balance |
Jobs are async: submit tools return a jobId, check tools poll it. Responses tell the agent exactly what to do next.
claude mcp add importly -e IMPORTLY_API_KEY=your_key -- npx -y importly-mcp
{
"mcpServers": {
"importly": {
"command": "npx",
"args": ["-y", "importly-mcp"],
"env": { "IMPORTLY_API_KEY": "your_key" }
}
}
}
(Claude Desktop: claude_desktop_config.json · Cursor: .cursor/mcp.json · Cline: MCP settings UI · Windsurf: ~/.codeium/windsurf/mcp_config.json)
Usage-based credits (downloads ~$0.01/MB, transcription per media minute) — see pricing. New accounts get $3.00 free.
MIT licensed. The server is a thin wrapper over the Importly API.