Piqt (iOS) – on device photo and video curator
I've just released Piqt, an on-device photo/video curator. For years I've wanted to get my Apple Photos library under control, but there just aren't enough hours in the day. I started out in Jupyter notebooks to see how much I can use AI models to automate it and Piqt is the result. My goals with Piqt: - Fully on device: I use a number of ML models, Apple Vision API, clustering algorithms to rank and sort your photos. Everything runs on device. - Private and Safe: Piqt has no accounts, no cloud services, no personal identifiers. Images or personal information never leave the device.…
What it does
Piqt curates every photo on your iPhone on-device, ranks your great and not-so-great photos, groups the near-identical shots, and files your screenshots. Explainable scores. Nothing uploaded. Nothing deleted without you.
Piqt scores every photo in your library on your iPhone then ranks it, groups the similar shots, and files the rest where you can find it. Easily find the pictures that matter to you. Nothing is uploaded. Nothing is deleted without you. Your phone dumps every selfie, milestone, portrait, experience, receipt, screenshot ... into one big bucket. It will not tell you which ones are good, and it will never make the other 9,999 worth opening. Your phone needs a curator. Curation is why people love museums: interpret a collection, organize it, and make it worth visiting. Photos gives you a vault and a search box. Piqt curates and gives you the collection. The plates on this page are illustrations…from piqt.app
In the maker’s words, at launch
I've just released Piqt, an on-device photo/video curator. For years I've wanted to get my Apple Photos library under control, but there just aren't enough hours in the day. I started out in Jupyter notebooks to see how much I can use AI models to automate it and Piqt is the result. My goals with Piqt: - Fully on device: I use a number of ML models, Apple Vision API, clustering algorithms to rank and sort your photos. Everything runs on device. - Private and Safe: Piqt has no accounts, no cloud services, no personal identifiers. Images or personal information never leave the device. Piqt also CAN'T delete your photos. I have a test and validation step in the release process to ensure no delete APIs exist in the code. Instead, Piqt stages things you want to clear out in an album, which you can delete when you're ready. - A library you can use. I didn't just want to clean out images or clear up space. There are plenty of apps for this. My ultimate goal is to make your (and my) photos easier to navigate I'm just looking for feedback on the app and the approach. I'd love to chat about the vision models I'm using and hear from you about any research or techniques I've missed. App Store Link: https://apps.apple.com/us/app/piqt-photo-video-curator/id677... Blog with detail on technical decisions: https://piqt.app/blog/piqt-design-values/
Does the same job
all alternatives →
- IVI've launched my iPhone journal app I've worked on for almost 2 years.2013 · hey.co · ▲139
Hi everyone - I've been working on this automatic journal for a long time. It takes your photos, makes them look nice, and also journals the places you go just by having your phone in your pocket. Would love your feedback!
Tag Your PhotosJul 2026 · apps.apple.com · ▲92AI-powered keywords for Apple Photos, 100% on your Mac
- WBWe built an appliance with Raspberry Pi that backs up from the cloud2012 · ▲17
http://picirus.com For Startup Weekend Tulsa, we created a small low-power appliance (built on Raspberry Pi) that you connect to your different cloud accounts -- Facebook, Flickr, Dropbox, etc. -- and it downloads a copy of all your stuff every night. I was thinking about all the people I love who share directly to the cloud, without a lot of planning for the case where their Facebook account gets hacked, their Dropbox account has a glitch, or one of these services they love goes out of business suddenly some day. We had a ton of fun building Picirus during Startup Weekend. I highly…
More ai this month
the category →
I trained a 125M-parameter transformer to autocomplete piano performances in real time (~108 notes/sec on an iPhone 15). The idea is basically GitHub Copilot or Tabnine, except instead of prompting it with code, you prompt it by playing a few notes on a MIDI piano. The model then continues what you played, entirely on-device. The app is free if anyone wants to try it. Happy to answer questions about the model, training, Core ML, or the many things that didn't work.
AI · 17d ago · simedw.com
Astute▲585Automate your B2B brand going viral, with new media creators
AI · 18d ago · company-app.joinastute.com
Hey HN, Henry from Cactus here! We previously released Cactus Needle, a 14MB agentic LLM for tool call, device use, and structured extraction for phones, wearables, smart homes, small robots and microcontrollers. We got really great feedback here, and have now incorporated the suggestions to release Needle 2. The whole model is a single 14MB binary that runs a full session in 28MB of RAM; 45m parameters at 2bit compression. Needle hits 500 tokens/sec decode speed on a Raspberry Pi 5, sits between 400-1,500 tokens/sec on VR devices like Meta Quest 3S and Apple Vision Pro, and ranges…
AI · 27d ago · cactuscompute.com
Launched alongside, August 2026
the whole month →- TL
Life & fun · 10d ago · louisabraham.github.io
- SA
Hello HN! I found that picking out plausible but diverse skin tones for my digital art and game development projects was kind of difficult, and I got curious about if there was a way to define a color space that made it easy. I've built a color picker and procedural generation algorithm based on the space as well as a bunch of other fun js features and demos throughout the page that use the equations. If you find it interesting, I have lots of explanations of how I built it and what properties the space has. The methodology might be a bit shaky, but hopefully the result is as helpful for…
Life & fun · Aug 2026 · toneyalexander.github.io
I trained a 125M-parameter transformer to autocomplete piano performances in real time (~108 notes/sec on an iPhone 15). The idea is basically GitHub Copilot or Tabnine, except instead of prompting it with code, you prompt it by playing a few notes on a MIDI piano. The model then continues what you played, entirely on-device. The app is free if anyone wants to try it. Happy to answer questions about the model, training, Core ML, or the many things that didn't work.
AI · 17d ago · simedw.com