signalHacker News Show HN2026-09-30
Show HN: Image Categorizer: sort a photo library with local vision models
Image Categorizer sorts photo folders using local AI via Ollama: a vision model describes each photo and flags accidental shots, a text model assigns categories refined in a browser report, and an exported script moves files or writes captions, all locally. On an RTX 4090, 1,195 photos took 80 minutes to describe and 8 minutes to sort with qwen3.6:35b and mistral-small3.2:24b.
- for who
- People with large photo libraries full of accidental shots, screenshots, and near-duplicates.
- why now
- New local AI photo sorter launches, offering private categorization without cloud uploads.
- what changes
- They can sort thousands of photos without manual review, and re-sort in minutes using saved descriptions, all offline.
- to do
- Install via uv and Ollama, pull the two models, and run the tool on a photo folder.
key points
- Uses Ollama with qwen3.6:35b vision and mistral-small3.2:24b text models
- 1,195 photos: 80 min describe, 8 min sort on an RTX 4090
- Exports bash script to move files or write captions, nothing deleted
#local ai#image categorization#photo organization#ollama#open source tool
score
score 8 out of 10. 0-10: how dense the facts are, multiplied by how much you can do with them after reading. 8+ means the topic's evidence bar is met: benchmarks and availability for a new model, amount and investors for a funding round, revenue figures for a solo-money story. Below 5 an item does not enter the digest. A press release scores 3 or less, a reprint loses 2, anything older than 14 days loses 1, a headline that misleads loses 3.
read the source