image-inputs

Image Inputs Example

This example demonstrates how to send images to vision models through the OpenRouter API.

Features Demonstrated

1. Single Image with URL - Send a publicly accessible image URL
2. Multiple Images - Send multiple images in a single request
3. Detail Levels - Control image analysis detail (low/high/auto)
4. ContentBuilder - Build complex messages with interleaved text and images
5. Base64 Encoding - Send local images using base64 encoding

Running the Example

# Set your API key
export OPENROUTER_API_KEY="your-api-key"

# Run the example
go run examples/image-inputs/main.go

Image Formats

OpenRouter supports the following image formats:

• PNG (image/png)
• JPEG (image/jpeg)
• WebP (image/webp)
• GIF (image/gif)

Image Sources

URL-based Images

Use publicly accessible image URLs. This is more efficient as it doesn't require local encoding:

message := openrouter.CreateUserMessageWithImage(
    "What's in this image?",
    "https://example.com/image.jpg",
)

Base64-encoded Images

For local files or private images:

// Automatically encodes the image to base64
message, err := openrouter.CreateUserMessageWithBase64Image(
    "What's in this image?",
    "path/to/image.jpg",
)

Detail Levels

Some models support detail level parameters:

• "low" - Faster and cheaper, suitable for general understanding
• "high" - More detailed analysis at higher cost
• "auto" - Let the model decide based on image size (default)

message := openrouter.CreateUserMessageWithImageDetail(
    "Describe this image in detail.",
    "https://example.com/image.jpg",
    "high",
)

Multiple Images

Send multiple images in a single request:

message := openrouter.CreateUserMessageWithImages(
    "Compare these images",
    "https://example.com/image1.jpg",
    "https://example.com/image2.jpg",
    "https://example.com/image3.jpg",
)

Content Builder

For complex messages with interleaved text and images:

content := openrouter.NewContentBuilder().
    AddText("Here's the first image:").
    AddImage("https://example.com/image1.jpg").
    AddText("And here's the second:").
    AddImageWithDetail("https://example.com/image2.jpg", "high").
    AddText("What are the differences?")

message := content.BuildMessage("user")

Model Support

Most modern vision models support image inputs, including:

• Google Gemini models (gemini-2.0-flash-thinking-exp, etc.)
• OpenAI GPT-4 Vision models
• Anthropic Claude 3 models
• And many others

Check the OpenRouter documentation for the latest list of supported models.

Important Notes

• The number of images you can send in a single request varies per provider and per model
• Image URLs must be publicly accessible
• Base64-encoded images increase request size, so use URLs when possible
• Some providers may have size limits on images
• Pricing may vary based on image size and detail level