README.md

---
title: Vibe Shopping
emoji: 🛍
colorFrom: green
colorTo: blue
sdk: gradio
sdk_version: 5.33.1
app_file: app.py
pinned: false
license: apache-2.0
short_description: AI powered clickless shopping experience with virtual try-on
tags:
  - agent-demo-track
---

This is embedding the actual app at https://huggingface.co/spaces/sitatech/vibe-shopping

# Vibe Shopping

**You may face cold-start which can take around 2-5 minutes, if you remain inactive for more than 5 minutes, your next interaction may also face cold-start. After a cold start you may need to reload the page, I will add**

Vibe Shopping is an AI-powered clickless shopping experience that allows users to virtually try on clothes and accessories.

It uses Qwen-2.5-VL-32B hosted on modal as the orchestrer, it as access to 3 MCP servers:

- [agora-mcp](https://github.com/Fewsats/agora-mcp)
- [fewsats-mcp](https://github.com/Fewsats/fewsats-mcp)
- [Vitrual Try-On MCP](./mcp_server.py) built as part of this project, can be found in the `mcp_server.py` file.

The virtual try on MCP uses [Flux-Fill](https://huggingface.co/black-forest-labs/FLUX.1-Fill-dev) combined with automatic masking, the model is hosted on modal.

Better docs comming soon with video.