[WordPress] 外掛分享: WearGraph – AI Virtual Try-On & Product Studio for WooCommerce

首頁 › 外掛目錄 › WearGraph – AI Virtual Try-On & Product Studio for WooCommerce
WordPress 外掛 WearGraph – AI Virtual Try-On & Product Studio for WooCommerce 的封面圖片
全新外掛
安裝啟用
—
尚無評分
2 天前
最後更新
—
問題解決
WordPress 6.5+ PHP 7.4+ v1.0.0 上架:2026-10-05

內容簡介

WearGraph 是一款專為 WooCommerce 設計的 AI 虛擬試衣間外掛,讓顧客能夠上傳自己的照片,預覽服裝在自己身上的效果,並可直接加入購物車。此外掛還提供產品編輯器中的產品工作室,幫助商家生成專業的產品圖片。

【主要功能】
• AI 虛擬試衣間,顧客可上傳照片預覽服裝效果
• 支援多種 AI 提供者,無需 WearGraph 帳號
• 產品工作室生成專業的產品圖片
• 自動添加試穿按鈕於可試穿產品
• 可隨時刪除上傳的照片
• 限制使用次數以控制 AI 費用

外掛標籤

開發者團隊

⬇ 下載最新版 (v1.0.0) 或搜尋安裝

① 下載 ZIP → 後台「外掛 › 安裝外掛 › 上傳外掛」
② 後台搜尋「WearGraph – AI Virtual Try-On & Product Studio for WooCommerce」→ 直接安裝(推薦)
📦 歷史版本下載

原文外掛簡介

Shoppers want to know how a garment will look on them, not only on a model. WearGraph adds an AI fitting room to your WooCommerce product pages: the shopper adds a photo, picks a colour or option, and gets an AI appearance preview of the product on themselves, with an Add to Cart button right beside it. For larger stores, the optional WearGraph Pro add-on adds brand kits, video, saved looks and more.
WearGraph also gives you a Product Studio inside the product editor, to turn your existing product photos into on-model and lifestyle images, and a Catalog screen that tells you which of your products are ready for virtual try-on and what to fix for the rest.
WearGraph runs inside your WordPress site with your own AI provider key. Choose Google Gemini, OpenAI, OpenRouter (one key for 50+ image models), Black Forest Labs FLUX or xAI Grok, connect one or several, and pick which provider makes product images, try-on previews and alt text. There is no WearGraph account, no subscription and no WearGraph server in between: you pay your provider directly for what you generate.

AI fitting room (virtual try-on)

A Try On button on Try-On Ready products: placed automatically, or with the Try On button block or the [weargraph_try_on] shortcode.
Upload a photo or use the camera, with on-screen tips for a good photo.
Shoppers must agree before any photo is uploaded. Location data and other hidden information are removed from the photo before it is stored.
Switch colours and options without uploading the photo again.
Add to Cart from the preview, into the shopper’s own cart.
Delete my photos at any time, straight from the fitting room.
Limits per visit, per connection and per day protect your AI bill, and a hook lets you add a CAPTCHA.
Preview mode: show the button only to store managers while you try it yourself.

Product Studio

Generate on-model, flat-lay and lifestyle images from your own product photos, with scenes, backgrounds, styles, reference images and “avoid” instructions.
Choose the aspect ratio and resolution (1K, 2K or 4K).
Save a result to the Media Library, set it as the product image, or add it to the gallery. Your original images are never overwritten.
Suggested alt text for product images.
A history of every image you created, with its settings and cost.

Catalog readiness

Checks every product’s type, category, attributes and images, without any AI and at no cost.
“N of M products are Try-On Ready”, with a plain reason for every product that is not.
Per-product and bulk overrides: switch try-on on or off, set the category, choose the source image.
Large catalogs are checked in small background batches.

Dashboard

Try-ons today, shoppers who tried on, add-to-cart, checkouts, orders and sales after try-on, images generated, estimated AI cost and failed generations.
A 14-day try-on chart, the most-tried products and recent activity.

Privacy by default

Shopper photos and previews are deleted when the visit ends (or after 24 hours, if you choose). An hourly clean-up enforces this.
Photos are private files served only to the shopper’s own browser, through links that expire after 10 minutes.
Logged-in customers’ fitting-room data is included in WordPress’s Export and Erase Personal Data tools.
Suggested privacy-policy text is added to Settings → Privacy → Policy guide.
A “Delete all shopper photos now” tool.

Cost control

A maximum cost per request and a daily spending cap (US$10 by default).
Failed and refused generations never count toward your limits.
Repeated identical requests reuse the earlier result instead of paying again.

WearGraph Pro
WearGraph Pro is a separate, optional add-on with features for larger stores (brand kits, video, saved looks and more). Everything described above works fully without it.
External services
WearGraph connects only to the AI provider(s) you choose in WearGraph → Settings → AI Provider, only after you have added your own API key for that provider, and only when a request runs. Providers without a saved key are never contacted. WearGraph itself has no server: it sends nothing to WearGraph or CodeSolz, and performs no license check, analytics or tracking.
What is sent, and when (to the provider set for that job):

When you click “Test connection” or “Refresh models”: your API key, sent to that provider to check the key or list its models. Gemini also receives the name of the image model and a short fixed text prompt (“Reply with the single word OK.”). Nothing is generated and no credits are used, except the few tokens of that Gemini check.
When you generate or edit an image in the Product Studio: your API key, the product images you selected, and the prompt built from your choices (scene, style, your own instructions).
When you ask for alt text: your API key, the product image and a prompt that includes the product name.
When a shopper requests a try-on preview: your API key, the shopper’s photo (after they agreed, with location and other hidden data removed), the product image, and a prompt describing the garment. No name, email address, IP address or account detail of the shopper is sent.

Nothing is sent when the catalog is checked, when pages load, or while no API key is saved.
To start a queued image or try-on preview without waiting for WP-Cron, WearGraph also sends a request to your own site (wp-admin/admin-ajax.php), signed so only WearGraph can trigger it. That request stays on your own server and is not an external service.
Google Gemini API (Google), at https://generativelanguage.googleapis.com/:

Gemini API Additional Terms of Service: https://ai.google.dev/gemini-api/terms
Google Privacy Policy: https://policies.google.com/privacy
Note: on Google’s unpaid tier, Google may use submitted content to improve its products, and human reviewers may read it. Use a key from a Google Cloud project with billing enabled.

OpenAI API (OpenAI), at https://api.openai.com/:

Services agreement: https://openai.com/policies/services-agreement/
Privacy policy: https://openai.com/policies/privacy-policy/

OpenRouter (OpenRouter, Inc.), at https://openrouter.ai/api/. OpenRouter passes the request to the company that runs the model you chose (for example Google, OpenAI, Black Forest Labs or ByteDance):

Terms of service: https://openrouter.ai/terms
Privacy policy: https://openrouter.ai/privacy

Black Forest Labs API (FLUX), at https://api.bfl.ai/. The finished image is downloaded right away from Black Forest Labs’ delivery servers (*.bfl.ai):

Terms of service: https://bfl.ai/legal/terms-of-service
Privacy policy: https://bfl.ai/legal/privacy-policy

xAI API (Grok Imagine), at https://api.x.ai/:

Terms of service: https://x.ai/legal/terms-of-service-enterprise
Privacy policy: https://x.ai/legal/privacy-policy

延伸相關外掛

文章
Filter
Mastodon