Frequently Asked Questions

Yes. Delle Model is a desktop app (Windows and Linux now, macOS coming soon) that runs gallery-dl for you — no command line needed. But it goes further than a typical wrapper: after gallery-dl downloads, Delle organizes everything into a browsable library with one profile per creator, plus tags, favorites, notes and a built-in viewer. Most gallery-dl GUIs only queue downloads; Delle is also what you use after the files land.

That is exactly Delle Model's niche. It uses yt-dlp (and gallery-dl) under the hood to fetch videos and galleries, then automatically sorts everything into per-creator profiles with tags, favorites and a player. Dedicated yt-dlp GUIs stop the moment the file is saved; Delle gives those files a home — a real library you can browse, search and deduplicate.

Point Delle Model at the folders where your media already lives and sync. Subfolders are auto-detected as creators, so you don't need to design a directory or output-template config — each creator becomes a profile holding all their videos and images. A folder watcher keeps the library in sync as new downloads arrive, so you skip the manual scripting most people do with gallery-dl's directory option or yt-dlp's output templates.

Yes. Delle Model reads the sidecar metadata gallery-dl and yt-dlp save next to your files and can backfill post dates, descriptions and tags back onto media you downloaded months ago. You then browse it all in a built-in viewer and swipe mode, grouped by creator — instead of digging through raw folders with no context.

Yes. Delle Model watches your media folders and auto-syncs files that are added, removed or modified, so new gallery-dl or yt-dlp downloads appear in the right creator's profile without a manual re-import. During sync it also hashes media for dedup and can backfill metadata from the sidecar files.

SHA-256 (exact) hashing finds byte-for-byte identical files — the same download saved twice. Perceptual hashing (Delle uses dHash) finds visually similar media, so it catches the same image or video saved in a different resolution, format or as a repost. Delle Model runs both automatically on sync, so duplicates and near-duplicates surface as part of your normal library workflow rather than as a separate cleanup chore.

Every image and video can take custom tags and be marked a favorite, and each creator profile holds notes and links. You can then filter and sort by tag, creator, date or media type, and bulk-select across pages to mass-tag, mass-move or recycle. A built-in Statistics dashboard totals your models, videos, images, favorites and runtime, and shows your most-viewed videos and most-used tags.

Yes. Stash is a self-hosted server you configure (Docker, ports, scrapers). Delle Model is a one-click desktop installer — no Docker, no server, no web setup. It also downloads content itself via gallery-dl and yt-dlp, which Stash does not, and it organizes around a creator-first profile (videos, images, tags, favorites, notes, links, timeline). If you wanted Stash without the setup friction, this is the lighter, local desktop option.

It is fully local. There is no cloud, no account and no telemetry. Your whole library lives on your disk as a portable file you can export with one click and restore on another machine. An internet connection is only used when you actively download content via gallery-dl or yt-dlp, or if you opt in to access your library over your own LAN.

Delle Model is a free closed beta. You get a free beta key by joining the Discord, which is also where new builds, support and the roadmap are posted. There is no payment and no subscription — download the Windows installer or the Linux AppImage and use a key from Discord to get in.