Skip to main content
The --image flag enables vision-based image analysis and description using models like GPT-4o.

​
Quick Start

praisonai "Describe this image" --image photo.png

​
Usage

​
Basic Image Description

praisonai "What's in this image?" --image /path/to/image.jpg
Expected Output: 
This image shows a scenic mountain landscape with snow-capped peaks 
reflecting in a crystal-clear lake. The foreground features pine trees 
and wildflowers, creating a classic alpine scene.

​
Detailed Analysis

praisonai "Analyze this image in detail, including colors, composition, and mood" --image photo.png

​
Multiple Images

praisonai "Compare these two images" --image image1.png,image2.png

​
With Custom Model

praisonai "Describe this image" --image photo.png --llm gpt-4o

​
Supported Formats

  • PNG (.png)
  • JPEG (.jpg, .jpeg)
  • GIF (.gif)
  • WebP (.webp)
  • BMP (.bmp)
  • SVG (.svg)
  • TIFF (.tiff)

​
Use Cases

​
Content Analysis

praisonai "What products are shown in this image?" --image product-photo.jpg

​
Accessibility

praisonai "Generate alt text for this image" --image website-banner.png

​
Document Analysis

praisonai "Extract text from this screenshot" --image screenshot.png

​
Code Review

praisonai "What does this diagram show?" --image architecture-diagram.png

​
Combine with Other Features

# With guardrail
praisonai "Describe this image" --image photo.png --guardrail "Keep description under 100 words"

# With metrics
praisonai "Analyze this chart" --image chart.png --metrics

# With save
praisonai "Describe this image" --image photo.png --save

​
Default Model

The default vision model is gpt-4o which provides excellent image understanding capabilities. You can override this with --llm: 
praisonai "Describe this" --image photo.png --llm gpt-4o-mini
Image description uses vision-capable models to analyze existing images. To generate new images from text, use --image-generate instead.