Image to Prompt Configuration
What Is This Feature?
Image to Prompt turns any image into a prompt for AI image generation.
For example, when you find an image online and want to create something similar with AI, hover over it to display the Convert to Prompt button. After you click the button, AI analyzes the image and returns the prompt that could have generated it in both Chinese and English.
What Do I Need?
Image analysis requires an AI service. You need to:
- Sign in to your account to use the official service
- Or configure a third-party API with your own API key
Where Do I Configure It?
Open the extension side panel → click the Settings icon at the bottom → select the My Account tab.
Under API Configuration, you can:
- Use the official service (sign-in required)
- Configure a third-party API by clicking Third-Party API Configuration
Option 1: Use the Official Service (Recommended)
Benefits
- No API key configuration
- A specialized vision model with strong results
- Available immediately after signing in
Steps
- Click Sign In on the My Account page
- After signing in, the official service card displays Professional Vision Model
- Click the activation button (✓ icon)
- You can use the service when Active appears
Note: The official service requires a Pro or Team subscription. Free users must upgrade.
Option 2: Configure a Third-Party API
If you have an API key, you can configure a third-party provider.
Open the Configuration Dialog
Click Third-Party API Configuration to open the configuration dialog.
Quick Configuration (Recommended)
The dialog opens on the Quick Configuration tab by default:
Step 1: Select a provider The menu groups providers by type:
- Chinese providers: Zhipu Qingyan, Alibaba Cloud Model Studio, Kimi, Xiaomi MiMo, Volcengine Ark
- International APIs: Anthropic Claude, OpenAI
- Aggregator: OpenRouter
Step 2: Select a model After selecting a provider, choose from its supported models.
Step 3: Enter an API key
- Paste your key into the API Key field
- Click Get API Key to open the provider's official website
Step 4: Save Click Save Configuration.
Custom Configuration
If your provider is not listed, open the Custom Configuration tab and enter:
- Configuration name: A name of your choice
- API format: Anthropic or OpenAI format
- API URL: The provider endpoint (HTTPS required)
- Model name: The model ID
- API Key: Your key
How to Use It
After completing the configuration, on any webpage:
- Hover over an image
- Wait a few seconds for the Convert to Prompt button to appear
- Click the button
- Wait for the AI analysis
- Receive prompts in Chinese and English
You can copy a prompt to another AI image-generation platform or save it to the extension's temporary prompt library.
Feature Toggle
The Image to Prompt toggle is at the bottom of the My Account page:
- On: The conversion button appears when you hover over an image
- Off: The button is hidden, but your configuration is retained
Switching Configurations
You can save multiple API configurations and switch at any time:
- Click the activation button (✓) on a configuration card to switch
- The active configuration is used for all image analysis
Pricing
- Official service: Usage quota is included with Pro and Team subscriptions
- Third-party API: The provider charges for usage; the extension does not
- Chinese providers often offer free quotas for evaluation
Security
Your API key is stored only in your browser. It is not uploaded to the cloud or synced, even when you are signed in.