YTS Movie Torrents

YTS is a popular movie torrent site known for high-quality releases with small file sizes. All movies include multiple quality options (720p, 1080p, 2160p/4K).

Site: https://yts.mx/ — API calls try this host first, then the yts.lt mirror unless you set YTS_API_BASE (no trailing slash; include /api/v2).

Reference

• Code Snippets: scripts/
• YTS API: search_movies.py tries yts.mx, then yts.lt, or set YTS_API_BASE
• qBittorrent Integration: See the qbittorrent skill

Python Package

• pyYify: Python library for the YTS JSON API (package name is historical; it works with YTS-compatible endpoints)

Available Scripts

Search Parameters

Usage

from pyYify import yify

movies = yify.search_movies("Inception", quality="1080p", minimum_rating=7)
for movie in movies:
    print(f"{movie.title} ({movie.year}) - IMDb: {movie.rating}")
    for torrent in movie.torrents:
        print(f"  {torrent.quality}: {torrent.size} - Seeds: {torrent.seeds}")

Direct API Access

import os
import requests

base = os.environ.get("YTS_API_BASE")  # e.g. https://yts.mx/api/v2
if not base:
    for candidate in ("https://yts.mx/api/v2", "https://yts.lt/api/v2"):
        try:
            r = requests.get(f"{candidate}/list_movies.json", params={"limit": 1}, timeout=15)
            r.json()
            base = candidate
            break
        except Exception:
            continue
if not base:
    base = "https://yts.mx/api/v2"
url = f"{base}/list_movies.json"
params = {"query_term": "Inception", "quality": "1080p", "sort_by": "seeds", "limit": 10}
response = requests.get(url, params=params)
data = response.json()

Integration with qBittorrent

from pyYify import yify
from skills.qbittorrent.scripts.qbittorrent_client import get_client, add_magnet

movies = yify.search_movies("Dune", quality="1080p")
if movies:
    torrent = next((t for t in movies[0].torrents if t.quality == "1080p"), movies[0].torrents[0])
    client = get_client()
    add_magnet(client, torrent.magnet)

Movie Selector Web App

A local web app (assets/movie_selector/index.html, served by movie_selector_server.py) for picking movies visually: a table with poster, screencap, plot summary, IMDb rating, and torrent links per quality. Hovering a poster shows an enlarged preview; screencaps are carousels with prev/next arrows over YTS's numbered stills; the list uses infinite scroll with the next page prefetched in the background.

🔥 Trending toggle: shows only the movies currently trending on yts.gg/trending-movies, in site order. The backend scrapes the trending page, resolves each movie's id from its page, and fetches full details (movie_details.json) with a 6-way thread pool - first load ~15-30s, then cached server-side for 10 minutes. Search/sort controls are disabled while trending is active; clicking Search or the toggle again returns to the normal newest-first list.

The browser queries the YTS API directly when it can (the API is CORS-enabled). When the YTS mirrors are unreachable from the browser's network (521s, ISP blocks), the page automatically falls back to the backend's /api/movies and /api/img endpoints, which fetch through the SSH SOCKS5 tunnel pool (CTFILE_PROXY_HOSTS in .env, ssh -D tunnels via the tvboxnow ssh_proxy_pool module). Proxied requests are restricted to allowlisted YTS hosts (yts.mx, yts.lt, yts.gg, movies-api.accel.li), with resolved-IP checks against loopback/private/reserved ranges before every outbound request. Selections persist in temp/yts_movie_selector/selection.json (override with YTS_SELECTION_PATH).

# 1. Start the app (default http://127.0.0.1:8010)
uv run python skills/yts/scripts/movie_selector_server.py
uv run python skills/yts/scripts/movie_selector_server.py --port 8010 --host 127.0.0.1

# 2. Open http://127.0.0.1:8011, search, and click a quality button to queue
#    exactly that torrent (e.g. "＋ 1080p · 1.85 GB"; click ✓ to remove)

# 3. Download everything queued - each entry uses the quality you queued
uv run python skills/yts/scripts/download_selected.py
uv run python skills/yts/scripts/download_selected.py --save-path /path/to/downloads
uv run python skills/yts/scripts/download_selected.py --list        # show queue
uv run python skills/yts/scripts/download_selected.py --dry-run     # print magnets only

Selection API

Selections keep their status (pending -> downloaded once handed to qBittorrent), so re-running download_selected.py only processes new picks. The agent workflow: start the server, let the user pick movies in the browser, then run download_selected.py - no per-movie CLI searching needed.

Genres

action, adventure, animation, biography, comedy, crime, documentary, drama, family, fantasy, history, horror, music, mystery, romance, sci-fi, sport, thriller, war, western