PlacePix
Open-source Self-Hosted Placeholder Image Service
Start new thread

PlacePix - Open-source Self-Hosted Placeholder Image Service

by•
PlacePix is a self‑hosted placeholder image server. Drop images into a folder, run Docker, and serve resized, filtered, and formatted images via simple URLs. Features: face‑aware cropping, filters (grayscale, blur, sepia, etc.), multiple formats (JPEG, PNG, WebP, AVIF), color/text placeholders, preset dimensions, watermark, srcset generation, upload API, S3 storage, and an admin dashboard. Self‑hosted · Open source · Zero config

Add a comment

Replies

Best
Maker
📌
🚀 Excited to launch PlacePix today! I built this because I kept needing placeholder images for prototypes, but didn't want to rely on third‑party services that might go down or limit my requests. With PlacePix, you own your image library and can serve it exactly how you need — resize, crop, filter, watermark, and more — all via simple URL parameters. A few things I'm particularly proud of: Smart face‑aware cropping using OpenCV — it automatically detects faces and centers them in the crop Preset dimensions for Instagram, YouTube, Facebook, TikTok, and more — no more guessing sizes SVG & gradient placeholder generators — perfect for wireframing and layouts Experimental AI generation with OVHcloud AI Endpoints integration Single Docker command to get started — docker run -d -p 3000:3000 -v ./images:/app/images riadvice/placepix:latest It's fully open‑source, self‑hosted, and works with S3‑compatible storage if you want to scale. I'd love to hear your feedback, feature requests, or use cases — what would make this even more useful for your workflow? Drop a comment below! 👇

Dropped a folder of images, ran the Docker compose file, and the face-aware cropping actually nailed it on a mixed bunch of group photos. The preset dimensions save me from typing query strings every time, which is a small thing but very satisfying.

Maker

 I am happy that you tried PlacePix and it saved you huge efforts. If you see that we can make it better just open a github issue. Did you try the "multiavatar" feature?