A powerful, privacy-first platform to organize, process, and stream your adult video collection - entirely on your own hardware.
Most video management tools are either too basic (a folder of files) or too invasive (cloud-hosted services that see everything). GoonHub fills the gap: a professional-grade video library that runs entirely on your machine, with features that rival commercial platforms.
- Zero cloud dependency - Your content never leaves your network
- Automatic processing - Thumbnails, sprite sheets, animated previews generated for you
- Smart search - Full-text search with 12+ filter types across your entire library
- Multi-user ready - Role-based access control for shared setups
Upload a video and GoonHub handles the rest. A multi-phase processing pipeline extracts metadata, generates thumbnails, creates seekbar previews, and builds animated previews - all in the background with real-time progress tracking.
- Metadata extraction - Duration, resolution, codecs, bitrate detected automatically via ffprobe
- Dual-resolution thumbnails - Small (320px) for grids, large (1280px) for detail views, both in WebP
- Sprite sheet generation - Grid of frames (configurable, default 96 frames) with VTT timing for seekbar hover preview
- Animated scene previews - Short MP4 clips sampling segments throughout the video, giving you a quick preview before watching
- Animated marker previews - Looping video clips at each bookmarked timestamp
- Configurable quality - CRF, resolution, frame intervals, WebP quality all tunable
- Non-blocking queue - DB-backed job queue handles 80,000+ videos without memory issues; workers process in parallel
Find exactly what you're looking for. Powered by Meilisearch, GoonHub delivers instant search results with filters that go far beyond basic text matching.
- Instant full-text search - Searches across titles, descriptions, filenames, actors, and tags
- 12+ filter types - Tags, actors, studio, duration range, date range, resolution, rating range, liked status, jizz count, marker labels, match type, and sort order
- Smart sort options - Newest, oldest, alphabetical, random (with shareable seed), most viewed, longest, highest rated
- Saved searches - Store your favorite filter combinations for one-click access
- URL-persisted filters - Share or bookmark any search configuration; browser back/forward works naturally
- User-specific filtering - Filter by your personal ratings, likes, and watch data
A full-featured video player with inline metadata editing, scene bookmarks, and interactive tracking - all on one page.
- Seekbar thumbnail preview - Hover over the timeline to see frame previews from generated sprite sheets
- Resume playback - Automatically remembers where you left off in every video
- Keyboard shortcuts - Space/K (play/pause), J/L (skip 10s), 0-9 (jump to %), F (fullscreen), T (theater mode), M (add marker)
- Inline metadata editing - Click any field to edit: title, description, release date, tags, actors, studio
- Star ratings - 1-10 scale with hover preview
- Like & jizz count - Personal interaction tracking per scene
- View counting - Automatic, deduplicated per user per 24h
- Related scenes - Score-weighted suggestions based on shared actors, tags, and studio
- Processing status - See the current processing state and reprocess if needed
Create timestamped bookmarks within any video. Each marker gets its own thumbnail - static or animated - and can be tagged and organized by label.
- One-click creation - Press M during playback to drop a marker at the current timestamp
- Custom labels & colors - Organize markers by category with color-coded labels
- Animated previews - Short looping video clips at each marker position (3-15 seconds, configurable)
- Tag markers individually - Apply tags to specific markers or entire label groups
- Dedicated markers hub - Browse all markers across your library, grouped by label or flat
- Click-to-jump - Click any marker to jump directly to that scene and timestamp
Curate collections of scenes with full playback control - reorder, shuffle, auto-advance, and track your progress.
- Drag-and-drop reordering - Arrange scenes in any order
- Auto-advance with countdown - Configurable countdown (5-30s) before the next scene plays
- Shuffle mode - Seeded randomization so shuffled order is reproducible and shareable
- Progress tracking - Resume the playlist from where you left off
- Visibility controls - Private, public, or unlisted playlists
- Create from search - Convert any search result into a playlist with one click
- Playlist sidebar - See your full playlist alongside the player, jump to any scene
Rich metadata management for performers, production studios, and custom tags - with dedicated browsing pages and scene associations.
- Actor profiles - Name, aliases, gender, birthday, physical attributes, career info, profile images
- Studio profiles - Name, URL, description, logo, parent network relationships
- Custom tags with colors - Create tags with custom color coding for visual organization
- Dedicated browse pages -
/actors,/studios,/markerseach with search, sort, and pagination - Scene associations - Link any scene to multiple actors, a studio, and unlimited tags
- Bulk association - Assign actors, studios, or tags to entire folders at once
Browse your video library by folder structure. Select multiple scenes and apply operations in bulk - tag, assign actors, match metadata, or delete.
- Folder navigation - Browse your storage paths like a file manager with breadcrumb navigation
- Multi-select - Ctrl+A to select all, Shift+Click for ranges, checkboxes for individual picks
- Bulk tagging - Apply tags to dozens of scenes at once
- Bulk actor/studio assignment - Set performers or studios for entire folders
- Bulk PornDB matching - Match multiple scenes against ThePornDB with confidence scoring and review before applying
- Bulk delete - Remove multiple scenes with a single confirmation
Automatically enrich your library with metadata from ThePornDB. Search for scene info, match performers, and pull in tags, descriptions, and images.
- Scene search - Query ThePornDB by filename or title to find metadata matches
- Performer lookup - Search and link actors from ThePornDB's database
- Auto-import metadata - Pull in title, description, tags, actors, studio, and thumbnail images
- Bulk matching - Match entire folders of scenes with progress tracking and confidence scores
- Review before applying - Preview matched metadata and cherry-pick what to import
Configure your landing page with the sections that matter to you. Mix and match recent uploads, watched scenes, liked content, tag-based collections, and more.
- Drag-to-reorder sections - Arrange your homepage layout exactly how you want it
- Section types - Recent uploads, recently watched, liked scenes, tag carousels, playlist collections, random picks
- Per-user configuration - Each user sees their own personalized homepage
- Section customization - Set the number of items and custom titles for each section
Track your viewing habits with a timeline view and activity chart. See what you watched, when, and pick up where you left off.
- Timeline view - Date-grouped list (Today, Yesterday, etc.) of watched scenes
- Activity chart - Visual bar chart showing daily watch activity
- Date range filtering - Last 7/30/90 days or custom range
- Resume from anywhere - Every scene remembers your last position
Watch your processing queue in real-time. The header badge shows active jobs, and a detailed popup breaks down progress by phase.
- Header indicator - Always-visible badge showing running, queued, and failed job counts
- Phase breakdown - See queue depth for metadata, thumbnail, sprites, and animated preview phases
- SSE-powered updates - No polling; updates stream in real-time via Server-Sent Events
- Auto-reconnection - Handles connection drops and resyncs automatically
Full control over processing, workers, scheduling, retries, and user management - all from the settings panel.
- Dynamic worker pools - Adjust worker count per phase at runtime, no restart needed
- Processing configuration - Tune thumbnail resolution, sprite grid size, preview quality, CRF values
- Scheduled triggers - Cron-based scheduling for automated processing (e.g., regenerate sprites nightly)
- Trigger types -
on_import,after_job,manual,scheduledwith cycle detection for dependencies - Retry configuration - Per-phase retry policies with exponential backoff
- Dead letter queue - Review and retry permanently failed jobs, or abandon them
- User management - Create users, assign roles, reset passwords, delete accounts
- RBAC - Fine-grained permissions (scenes:upload, scenes:trash, etc.) enforced at API level
- Trash management - View soft-deleted scenes, restore or permanently purge
- Storage paths - Add, edit, or remove video directories; trigger scans for new files
- Manual bulk processing - Reprocess metadata, thumbnails, sprites, or previews for all scenes or only missing ones
Point GoonHub at your video directories and it automatically discovers new files, detects moved files, and avoids duplicates.
- Automatic discovery - Scans configured storage paths for new video files (MP4, MKV, AVI, MOV, WEBM, WMV, M4V)
- Move detection - Recognizes files that have been moved between folders (via size + filename matching) to avoid duplicates
- Batch creation - Efficient batch processing (50 scenes per DB flush)
- Real-time progress - SSE-powered progress updates during scans
- Orphan detection - Automatically trashes scenes whose files are no longer found
Built for privacy from the ground up, with proper authentication, authorization, and account protection.
- PASETO tokens - Modern, secure token format (no JWT vulnerabilities)
- Role-based access control - Admin and user roles with fine-grained permissions
- Account lockout - Automatic lockout after failed login attempts (configurable threshold and duration)
- Login rate limiting - Per-IP rate limiting on authentication endpoints
- Token revocation - Proper logout with server-side token invalidation
- Cross-tab sync - Logout propagates across all open browser tabs
- Security headers - X-Frame-Options, CSP, HSTS, X-Content-Type-Options
| Layer | Technology |
|---|---|
| Backend | Go 1.24, Gin, GORM |
| Frontend | Nuxt 4, Vue 3, Tailwind CSS 4, Pinia 3 |
| Database | PostgreSQL 18 |
| Search | Meilisearch v1.33 |
| Video Processing | ffmpeg / ffprobe |
| Auth | PASETO v2 symmetric |
| Player | video.js 8, media-chrome 4 |
| DI | Google Wire |
| Deployment | Docker, multi-arch (amd64/arm64) |
git clone https://github.com/your-org/goonhub.git
cd goonhub/docker
# Set your secrets
cp .env.example .env
# Edit .env with your passwords and keys
# Start everything
docker compose -f docker-compose.prod.yml up -dGoonHub will be available at http://localhost:8080. An admin account is created automatically on first startup.
# Prerequisites: Go 1.24+, Bun, ffmpeg, PostgreSQL, Meilisearch
# Frontend build
cd web && bun install && bun run generate && cd ..
# Backend build
go build -o goonhub ./cmd/server
# Run
GOONHUB_CONFIG=config.yaml ./goonhubAll settings are configured via YAML file or environment variables:
# Via config file
GOONHUB_CONFIG=/path/to/config.yaml ./goonhub
# Via environment variables (dots become underscores)
GOONHUB_SERVER_PORT=8080 \
GOONHUB_DATABASE_PASSWORD=secret \
GOONHUB_AUTH_PASETO_SECRET=$(openssl rand -base64 32 | head -c 32) \
./goonhubSee docker/config.prod.yaml for all available configuration options.
%%{ init : { "theme" : "default" }}%%
graph TD
%% Top Layer
Nuxt["Nuxt 4 SPA<br>(embedded in Go)"]
%% Middle Layer
Gin["Gin HTTP API<br>Auth | RBAC | SSE"]
%% Service Layer
Core["Core Services<br>Scene, Auth,<br>Search, Jobs"]
Workers["Worker Pools<br>Metadata, Thumbnails,<br>Sprites, Previews"]
Meili["Meilisearch<br>Full-text<br>search index"]
%% Storage Layer
Postgres[("PostgreSQL 18<br>Scenes, Users, Jobs, Settings,<br>Markers, Playlists, History")]
%% Relationships
Nuxt --> Gin
Gin --> Core
Gin --> Workers
Gin --> Meili
Core --> Postgres
Workers --> Postgres
Key architectural decisions:
- DB-backed job queue - PostgreSQL acts as an infinite-capacity queue; bounded worker channels prevent memory explosion. Handles 80,000+ videos without issue.
- Embedded frontend - The Nuxt SPA is compiled and embedded into the Go binary via
embed.FSfor true single-binary deployment. - Event-driven real-time - An internal EventBus publishes events to SSE streams. No polling.
- Graceful shutdown - Lifecycle manager tracks all goroutines and coordinates shutdown with configurable timeout.
# Start dependencies
cd docker && docker compose up -d
# Backend (port 8080)
GOONHUB_CONFIG=config-dev.yaml go run ./cmd/server
# Frontend dev server (port 3000, proxies API to backend)
cd web && bun run dev
# Run tests
make test-race
# Regenerate mocks after interface changes
make mocks
# Regenerate Wire DI after provider changes
go run github.com/google/wire/cmd/wire ./internal/wire