image2prompt is a lightweight tool designed for creators, designers, and AI users, Used to generate Prompt words for any image on the web page with one click.
When you hover over an image on any webpage, a small icon button will appear in the bottom-right corner 👇
Click it, and the extension will automatically upload the image, generate a prompt, copy it to your clipboard, and open your preferred AI platform to reproduce the same style.
| Feature Module | Description |
|---|---|
| 🧩 Model Selection | Switch between Gemini 2.5 Flash and Zhipu GLM-4V — each keeps its own API key and model ID |
| 🌏 Multilingual Output | Generate prompts in 20 different languages |
| 🖼️ Image Size Filter | Only displays the button for images larger than the configured size (default: 256×256) |
| 📒 Generation History | View all your generated prompts anytime |
| 🎨 Custom Platform Jump | Configure the default AI platform: OpenAI / Gemini / StableDiffusion / JiMeng / Keling / Doubao / Hailuo AI / Custom |
| 💬 Prompt Templates | Edit and customize prompt generation templates to build your unique style |
| ✍️ Custom Instructions | Enable a pre-generation dialog so you can blend extra guidance into every output |
| 🧭 Internationalized UI | Easily switch between English and Chinese |
| 🪶 Lightweight UI | Inspired by shadcnUI, built with custom-drawn components and no third-party dependencies |
| 🔓 Open Source & Free | 100% open-source and completely free to use |
| 🧮 Aspect Ratio Presets | Pick cinematic ratios or add your own so every prompt respects the frame you need |
| 🚫 Domain Filters | Hide the capture button on sites you exclude so browsing stays distraction-free |
| 🖼️ Local Uploads | Drop local images directly in the settings page and generate prompts without leaving the dashboard |
- Clone or download this repository:
git clone https://github.com/pingan8787/image2prompt.git- Install the extension
After downloading the project, open the Chrome/Edge Extensions page: chrome://extensions/ / edge://extensions/ enable Developer mode, then drag the entire project folder into the page.
Alternatively, click Load unpacked, then select the project folder.
After installing the extension, open the configuration page to choose your provider and set the corresponding API key:
- Gemini: Create a Gemini API key
- Zhipu AI: Zhipu model overview & console
Then, whenever you hover over an image on any webpage, you’ll see an “Edit” button at the bottom-right corner — click it to generate a prompt!
Want to blend in extra guidance (for example, “switch the background to a neon city”)? Enable Custom instructions input in Settings → Prompt Generation to open a dialog before each run and merge your tweaks with the system prompt.
Scan the WeChat QR code below to get in touch:
