Quick Start
Set up Pindeck and upload your first image in minutes
AI Analysis
Automatic title, tags, color palette, and category extraction
AI Variations
Generate cinematic shot variations using fal.ai
Boards
Organize images into boards and export as pitch decks
Discord Integration
Import images from Discord with moderation queue
API Reference
Explore the full Convex function and HTTP endpoint reference
What Pindeck does
Upload your reference images
Drag and drop images through the Upload tab. Pindeck stores them in Convex (with automatic NextCloud persistence) and immediately queues them for AI analysis.
AI analyzes every image automatically
A Vision Language Model (via OpenRouter) extracts a title, description, 5–10 tags, a 5-color palette, visual style, category, and project group for every image — no manual tagging required.
Generate cinematic variations
Choose a modification mode (shot variation, b-roll, action shot, style variation, and more) and Pindeck generates new images using fal.ai Nano Banana Pro, linked back to the original via parent-child lineage.
Key capabilities
- Gallery view — Responsive masonry grid with search, category/group filters, like and view counters, and a full detail modal
- Table view — Sortable, filterable data table with global search, tag/color filtering, and pagination
- Project rows — Group images by project name for shot-by-shot organization
- Discord bot — Import images via emoji reaction or slash command, with an approval queue before AI processing
- NextCloud storage — Originals, previews, and multi-resolution derivatives (320 px, 768 px, 1280 px) stored on your own infrastructure
- Real-time updates — Convex subscriptions keep the UI in sync without manual refreshes
Pindeck is a production-first application. All scripts and build commands enforce a connection to your production Convex deployment before running.