ListingForge: a local Gemma tool that drafts shop listings from product photos

Description and use

ListingForge is a local‑first tool that drafts shop listings from product photos using Gemma 3 4B, keeping all data on the seller’s machine.

Problem

Small online sellers find writing product listings tedious and repetitive, especially when they only have product photos.

Outcome

ListingForge produces a draft listing that reflects only what is visible in the photos, flags mixed products, and prompts the seller for missing information, all while keeping data local and incurring no per‑listing cost.

Workflow

• 1. Upload 1–4 product photos in the browser; the client resizes them to 1024px JPEG. 2. The local Node.js server sends the images to Gemma 3 4B via Ollama. 3. Gemma performs a same‑product check; if photos are mixed, the tool lists each product separately. 4. Gemma drafts title, description, tags, and condition note, avoiding prohibited words. 5. The server adds a to‑do list for size, material, price, shipping, and handmade/vintage/mass‑produced status. 6. The seller reviews the draft, fills in missing details, and publishes the listing.

Evidence

ListingForge helps a small online seller turn photos of one product into a first draft of a shop listing.

Source

It runs entirely on your own computer using a local Ollama server and the Gemma 3 4B vision model.

Source

When you select or drop up to 4 product photos, the browser resizes them client‑side using an HTML canvas to a maximum of 1024px on the longest side (JPEG at 0.8 quality).

Source

Gemma wrote one listing about two of them and dropped the other two. It also used words like "handmade", "hand‑thrown" and "luxurious", which a photo can't prove.

Source

A separate yes/no check ("do all photos show the same product?") before any listing is written. With four different products it now names them and stops.

Source

It never generates prices or sizes.

Source