Image Input (Vision)

Some AI models available in Snowgoose have vision capabilities, meaning they can “see” and analyze images you provide.

How to Use Image Input

1. Select a Vision-Capable Model: Choose a model that supports vision (e.g., GPT-4o, Claude 3 models, Gemini models) from the Model Selector.
2. Upload an Image: When a vision model is selected, an Image Upload button (image icon) will appear next to the text input area at the bottom.
   • Click the image icon.
   • Select an image file from your device.
   • The name of the selected file will appear below the input area. You can click the ‘x’ next to the filename to remove it before sending.
3. Add Your Prompt: Type a message in the text input area asking the AI to do something with the image (e.g., “Describe this image,” “What text is visible in this picture?”, “Identify the objects in this photo.”).
4. Send: Click the Send button (arrow_upward icon).

The AI will process both your text prompt and the uploaded image to generate its response.

Use Cases

• Image Description: Get detailed descriptions of what’s in an image.
• Text Extraction (OCR): Read text from signs, documents, or screenshots.
• Object Recognition: Identify objects, landmarks, or people (use responsibly).
• Data Analysis: Analyze charts, graphs, or diagrams within images.
• Troubleshooting: Upload screenshots of error messages or UI problems for analysis.

Note: Image uploads are associated with the single message you send them with. They don’t persist across multiple messages in the chat.