liangdabiao/ecom-details-image-ui — explained in plain English
Analysis updated 2026-05-18
Generate a white-background hero image for a product from a short text description.
Upload a product photo and restyle its background, pose, or lighting while keeping the product itself unchanged.
Produce social-media style lifestyle images for platforms like Xiaohongshu from a product description.
Get an AI-written image generation prompt to use in another tool without generating an image here.
| liangdabiao/ecom-details-image-ui | alexmt/mobile-for-argocd | alvinunreal/juice | |
|---|---|---|---|
| Stars | 32 | 32 | 32 |
| Language | TypeScript | TypeScript | TypeScript |
| Setup difficulty | moderate | moderate | easy |
| Complexity | 3/5 | 3/5 | 3/5 |
| Audience | pm founder | ops devops | developer |
Figures from each repo's GitHub metadata at analysis time.
Deploying your own instance requires API keys for both the AI Gateway chat model and the Agnes AI image service.
EcomImageChat is a web app for generating professional-looking e-commerce product photos using AI, aimed at sellers who need product images fast. A user types a short description of their product in Chinese, such as "blue bluetooth earbuds" or "beige ceramic coffee cup," along with what kind of image they want, like a plain white background hero shot or a moody lifestyle photo styled for social platforms. The AI picks from 25 built in e-commerce scene templates and writes an image generation prompt automatically, then returns a finished image in about 15 to 30 seconds. Users can also upload a reference photo of their actual product. When they do, the app forces an image to image workflow that keeps the product's real color, material, and print intact while only changing the background, pose, lighting, or overall style around it. There is also a prompt only mode: if a message doesn't include a word like "generate" or "produce," the AI just writes out an English prompt and description the user can copy into another tool instead of generating an image itself. Under the hood, the app runs on Tencent Cloud's EdgeOne Makers platform and is built on the Claude Agent SDK. The main chat model is DeepSeek, accessed through an Anthropic-compatible protocol, while image recognition and image generation both go through a separate provider called Agnes AI. Uploaded and generated images are stored in EdgeOne's blob storage, and progress updates during upload, recognition, and generation are streamed to the browser in real time using server-sent events. Each chat session is limited to producing one image within a 120 second window to keep response times predictable. The frontend is built with React, Tailwind CSS, and Vite. Developers can deploy their own copy either through a one click deploy button to EdgeOne Makers, or by cloning the repository and using the EdgeOne CLI to develop locally and deploy manually. Setting it up requires API keys for the AI Gateway chat model and for the Agnes AI image service, configured through the EdgeOne Makers console rather than the command line.
A web app that turns a short product description or reference photo into a professional e-commerce product image using AI, in under 30 seconds.
Mainly TypeScript. The stack also includes TypeScript, React, Tailwind CSS.
Setup difficulty is rated moderate, with roughly 30min to a first successful run.
Mainly pm founder.
This repo across BitVibe Labs
double-check against the repo, no cap.