Open Source Torrent Aggregator

Find Any Torrent,
From Every Source

Search across YTS, EZTV, Knaben, and Prowlarr. Quality-scored, deduplicated, and enriched with TMDB metadata.

Try:

0+Movies

0+TV Shows

0+Torrents

0+Active Seeders

How It Works

A complete pipeline from torrent discovery to enriched, searchable content.

Multi-Source Ingestion

Aggregates torrents from YTS, EZTV, Knaben, and Prowlarr into a single searchable index.

Quality Scoring

Every torrent scored 0–100 based on resolution, source, codec, seeders, and file size.

TMDB Enrichment

Automatic metadata enrichment with posters, descriptions, genres, and translations in 11 languages.

Smart Deduplication

Uses info_hash as a universal key to eliminate duplicates across all sources.

Full-Text Search

Search across titles in multiple languages with PostgreSQL full-text search and GIN indexes.

Scheduled Updates

Built-in cron scheduler keeps the index fresh with configurable update intervals.

MCP Integration

Ask Any AI to Find Torrents

Connect via the Model Context Protocol. Search, filter, and download using natural language from your favorite AI assistant.

AI Assistant

Download The Matrix in 1080p

search_torrents({ q: "The Matrix", quality: "1080p" }) Found The Matrix (1999) — IMDb 8.7

Best match

1080p · x2652.1 GB847 seedersScore 92

Ask anything about movies or TV shows...

Works with any MCP-compatible client

Claude

ChatGPT

OpenClaw

Cursor

Windsurf

Gemini

Copilot

Grok

REST API + MCP

Built for LLMs & Developers

JSON API and MCP server ready for AI agents, LLMs, and traditional apps. Full-text search, smart filtering, and quality-scored results — accessible via HTTP or the Model Context Protocol.

REST API

Full-text search with multi-language support
Filter by type, genre, year, quality, and rating
Sort by relevance, seeders, rating, or date
Quality-scored torrents with magnet URLs

MCP Servernpx torrent-api-mcp

Connect from Claude, Cursor, Windsurf, or any MCP-compatible client. Let AI agents search torrents and access metadata using natural language.

Terminal

// claude_desktop_config.json
{
  "mcpServers": {
    "torrent-api": {
      "command": "npx",
      "args": ["torrent-api-mcp"]
    }
  }
}

// "Search torrents for Dune in 4K"
→ search_torrents({ q: "Dune", quality: "2160p" })

// "What sources are available?"
→ get_stats()

// "Find top-seeded sci-fi movies from 2024"
→ search_torrents({
    genre: "sci-fi",
    year_min: 2024,
    sort: "seeders"
  })

Find Any Torrent,From Every Source