I built a listing writer that runs on the seller's own laptop — open-weight models, zero cloud

Описание и применение

Разработан локальный генератор описаний товаров, который работает без облака и не требует оплаты за каждое объявление, обеспечивая конфиденциальность данных продавца.

Задача

A reseller needs to write dozens of listings manually, incurring per-listing costs and exposing sensitive data to cloud services.

Результат

Listing Studio produced a 56‑character Etsy title, 12 tags (clipped to 20 chars), a GBP 25–40 price band, and a fallback description in 49 s on CPU. A second run yielded a shorter title and a lower price band, demonstrating the model’s drafting role.

Рабочий процесс

• Describe an item once in a form
• Python builds a prompt with platform rules
• Ollama runs an open‑weight model
• Model outputs prose
• Python clamps titles, normalises tags, caps price band
• Results are exported as listings and CSV
• Optional history stored in SQLite

Доказательства

A reseller photographs forty things a week and then writes forty listings by hand.

Источник

Listing Studio: describe an item once and it writes a platform-shaped title, description, tags and price range for Etsy, eBay, Depop or Vinted — running entirely on your own machine.

Источник

The model writes prose, Python decides every hard limit.

Источник

First run, ~49 seconds on CPU, unedited: Title: Vintage Weathered Pine Wood Birdhouse by Local Carpenter (56/140 chars) … Price: GBP 25–40

Источник

I wanted the photo to be the input — point the camera at the item, get the listing. It is wired up, and it is not the default, for reasons worth sharing…

Источник