ListSmart: Turn Product Photos into Meesho/Etsy Listings Locally with AI

Description and use

ListSmart automatically generates Meesho and Etsy listings from product photos using local Gemma and Qwen3‑8B models, with SerpApi for pricing. It saves time and eliminates per‑listing API costs.

Problem

Small sellers need to create product listings quickly, accurately, and without incurring per-listing API costs.

Outcome

ListSmart reduces listing creation time, improves consistency, and eliminates per‑listing API costs by performing inference locally.

Workflow

• Capture product photo
• Run local vision model (Gemma) to extract details
• Fine‑tuned Qwen3‑8B generates listing text
• Query SerpApi for competitor prices
• Display editable listing with counters and copy button
• Seller reviews and publishes
• Optional subscription for additional listings

Evidence

ListSmart converts product photos into ready-to-publish Meesho and Etsy listings, including titles, descriptions, five bullet points, search tags, and suggested prices.

Source

The project uses Gemma for vision, Qwen3-8B with LoRA fine-tuning for listing generation, Tinker for training, and SerpApi for competitor pricing.

Source

My first fine‑tune reduced validation loss from 2.61 to 1.48, but performed worse than the base model on important quality metrics.

Source

The blind evaluation preferred the base model 77.5% of the time, compared with just 0.8% for v1.

Source