Fridge to List: Snap Your Fridge, Get a Grocery List (Gemma 3 + Ollama, All Local)

Description and use

Fridge to List automatically generates a grocery list from fridge photos using a local Gemma 3 4B vision model via Ollama, ensuring privacy and no API costs.

Problem

Users need a quick way to generate a grocery list from fridge photos without manual inspection.

Outcome

The app generated a grocery list in 7.5 seconds from two photos on a MacBook Air, eliminating the need for manual inspection.

Workflow

• Upload 1–4 fridge photos via browser UI
• Resize photos in browser and send to local Node.js server
• Server forwards images to Ollama running Gemma 3 4B
• Model returns JSON with visible items, quantities, confidence, and a "could not tell" list
• App compares detected items against user’s staple list
• App displays editable grocery list and a panel showing model’s detections with confidence

Evidence

Fridge to List turns photos of your fridge into a grocery list, so you don't have to stand there with the door open writing things down.

Source

The core is Gemma 3 (4B), an open-weight vision model, served by Ollama on my own MacBook Air.

Source

On my MacBook Air, the app reported 7.5 seconds to write the list from two photos (Gemma 3 4B, locally).

Source