View as Markdown Open in ChatGPT Open in Claude

ImageModel Interface

Interface that all image generation model implementations must satisfy.

Interface Definition

type ImageModel interface {
    // Metadata
    SpecificationVersion() string
    Provider() string
    ModelID() string

    // Image generation
    DoGenerate(ctx context.Context, opts *ImageGenerateOptions) (*types.ImageResult, error)
}

Methods

Method	Parameters	Returns	Description
SpecificationVersion()	-	string	Specification version
Provider()	-	string	Provider name
ModelID()	-	string	Model identifier
DoGenerate()	ctx, *ImageGenerateOptions	*types.ImageResult, error	Generate images

Optional Capability Methods

A model may optionally implement these methods to advertise whether it consumes ImageGenerateOptions.Files / .Mask for image editing. They are not part of the ImageModel interface itself, so implementing them is optional; callers read them through provider.ImageModelSupportsFileInputs / provider.ImageModelSupportsMaskInputs, which return nil (unknown) for a model that doesn't implement the method:

SupportsFileInputs() *bool // nil means unknown; only route file-input edits when this is non-nil true
SupportsMaskInputs() *bool // nil means unknown; advertised separately since some models accept files without masks

middleware.WrapImageModel preserves these from the wrapped model and lets an ImageModelMiddleware override either one via OverrideSupportsFileInputs / OverrideSupportsMaskInputs.

ImageGenerateOptions

type ImageGenerateOptions struct {
    Prompt  string
    N       *int
    Size    string
    Quality string
    Style   string
}

Examples

Using an Image Model

package main

import (
    "context"
    "log"
    "os"

    "github.com/digitallysavvy/go-ai/pkg/provider"
    "github.com/digitallysavvy/go-ai/pkg/providers/openai"
)

func main() {
    p := openai.New(openai.Config{
        APIKey: "your-api-key",
    })
    model, err := p.ImageModel("dall-e-3")
    if err != nil {
        log.Fatal(err)
    }

    result, err := model.DoGenerate(context.Background(), &provider.ImageGenerateOptions{
        Prompt: "A serene mountain landscape",
        Size:   "1024x1024",
    })
    if err != nil {
        log.Fatal(err)
    }

    // Save first image
    if len(result.Images) > 0 {
        os.WriteFile("output.png", result.Images[0], 0644)
    }
}

Interface Definition​

Methods​

Optional Capability Methods​

ImageGenerateOptions​

Examples​

Using an Image Model​

See Also​