Generate Images

Overview

Generates images from text prompts using pure image generation models (DALL-E, GPT Image). For multimodal models like gemini-2.5-flash-image, use the /v1/chat/completions endpoint instead.

Method Signature

func (r *ImageService) Generate(
    ctx context.Context,
    body ImageGenerateParams,
    opts ...option.RequestOption,
) (*ImagesResponse, error)

Request Parameters

prompt

string

required

Text description of the desired image(s). Maximum length:

32,000 characters for gpt-image-1
1,000 characters for dall-e-2
4,000 characters for dall-e-3

model

string

default:"openai/dall-e-2"

The model to use for image generation:

openai/dall-e-2 - Fast, cost-effective
openai/dall-e-3 - Higher quality, more detailed
openai/gpt-image-1 - Latest model with advanced features

int64

default:"1"

Number of images to generate. Must be between 1 and 10. Note: dall-e-3 only supports n=1.

size

string

Size of generated images:

For dall-e-2: 256x256, 512x512, 1024x1024
For dall-e-3: 1024x1024, 1792x1024, 1024x1792
For gpt-image-1: 1024x1024, 1536x1024, 1024x1536, auto (default)

quality

string

default:"auto"

Quality of the generated image:

auto - Automatically select best quality
For gpt-image-1: high, medium, low
For dall-e-3: hd, standard
For dall-e-2: standard only

style

string

Style of the generated images (dall-e-3 only):

vivid - Hyper-real and dramatic images
natural - More natural, less hyper-real images

response_format

string

default:"url"

Format for returned images (dall-e-2 and dall-e-3):

url - Returns URLs (valid for 60 minutes)
b64_json - Returns base64-encoded JSON

Note: gpt-image-1 always returns base64-encoded images

output_format

string

default:"png"

Output image format (gpt-image-1 only):

png - PNG format
jpeg - JPEG format
webp - WebP format

output_compression

int64

default:"100"

Compression level 0-100% (gpt-image-1 with webp or jpeg only)

background

string

default:"auto"

Background transparency (gpt-image-1 only):

auto - Model determines best background
transparent - Transparent background (requires png or webp)
opaque - Opaque background

moderation

string

default:"auto"

Content moderation level (gpt-image-1 only):

auto - Standard filtering
low - Less restrictive filtering

stream

bool

default:"false"

Generate in streaming mode (gpt-image-1 only)

partial_images

int64

default:"0"

Number of partial images for streaming (0-3, gpt-image-1 only)

user

string

Unique identifier for end-user for abuse monitoring

Response Fields

created

int64

required

Unix timestamp when images were created

data

[]Image

required

Array of generated images

Show Image fields

url

string

URL of the generated image (if response_format=url)

b64_json

string

Base64-encoded image data (if response_format=b64_json)

revised_prompt

string

Revised prompt used for generation (dall-e-3 only)

Code Examples

Basic Image Generation

package main

import (
    "context"
    "fmt"
    "log"

    dedalus "github.com/dedalus-labs/dedalus-sdk-go"
    "github.com/dedalus-labs/dedalus-sdk-go/option"
)

func main() {
    client := dedalus.NewClient(
        option.WithAPIKey("your-api-key"),
    )

    ctx := context.Background()
    
    response, err := client.Images.Generate(ctx, dedalus.ImageGenerateParams{
        CreateImageRequest: dedalus.CreateImageRequestParam{
            Prompt: dedalus.F("A serene landscape with mountains and a lake at sunset"),
            Model:  dedalus.F("openai/dall-e-3"),
            Size:   dedalus.F(dedalus.CreateImageRequestSize1024x1024),
        },
    })

    if err != nil {
        log.Fatal(err)
    }

    fmt.Println("Image URL:", response.Data[0].URL)
    if response.Data[0].RevisedPrompt != "" {
        fmt.Println("Revised Prompt:", response.Data[0].RevisedPrompt)
    }
}

Generate Multiple Images

response, err := client.Images.Generate(ctx, dedalus.ImageGenerateParams{
    CreateImageRequest: dedalus.CreateImageRequestParam{
        Prompt: dedalus.F("A cute robot playing with a puppy"),
        Model:  dedalus.F("openai/dall-e-2"),
        N:      dedalus.F(int64(4)),
        Size:   dedalus.F(dedalus.CreateImageRequestSize512x512),
    },
})

if err != nil {
    log.Fatal(err)
}

for i, image := range response.Data {
    fmt.Printf("Image %d: %s\n", i+1, image.URL)
}

High Quality with Specific Style

response, err := client.Images.Generate(ctx, dedalus.ImageGenerateParams{
    CreateImageRequest: dedalus.CreateImageRequestParam{
        Prompt:  dedalus.F("A futuristic city with flying cars and neon lights"),
        Model:   dedalus.F("openai/dall-e-3"),
        Size:    dedalus.F(dedalus.CreateImageRequestSize1792x1024),
        Quality: dedalus.F(dedalus.CreateImageRequestQualityHD),
        Style:   dedalus.F(dedalus.CreateImageRequestStyleVivid),
    },
})

Base64 Encoded Image

import (
    "encoding/base64"
    "os"
)

response, err := client.Images.Generate(ctx, dedalus.ImageGenerateParams{
    CreateImageRequest: dedalus.CreateImageRequestParam{
        Prompt:         dedalus.F("A watercolor painting of a garden"),
        Model:          dedalus.F("openai/dall-e-2"),
        ResponseFormat: dedalus.F(dedalus.CreateImageRequestResponseFormatB64JSON),
    },
})

if err != nil {
    log.Fatal(err)
}

// Decode and save image
imageData, err := base64.StdEncoding.DecodeString(response.Data[0].B64JSON)
if err != nil {
    log.Fatal(err)
}

err = os.WriteFile("generated_image.png", imageData, 0644)
if err != nil {
    log.Fatal(err)
}

GPT Image 1 with Advanced Features

response, err := client.Images.Generate(ctx, dedalus.ImageGenerateParams{
    CreateImageRequest: dedalus.CreateImageRequestParam{
        Prompt:            dedalus.F("A product photo of a sleek smartphone on a clean white background"),
        Model:             dedalus.F("openai/gpt-image-1"),
        Size:              dedalus.F(dedalus.CreateImageRequestSizeAuto),
        Quality:           dedalus.F(dedalus.CreateImageRequestQualityHigh),
        OutputFormat:      dedalus.F(dedalus.CreateImageRequestOutputFormatWebp),
        OutputCompression: dedalus.F(int64(85)),
        Background:        dedalus.F(dedalus.CreateImageRequestBackgroundOpaque),
    },
})

Transparent Background

response, err := client.Images.Generate(ctx, dedalus.ImageGenerateParams{
    CreateImageRequest: dedalus.CreateImageRequestParam{
        Prompt:       dedalus.F("A cartoon character, centered, no background"),
        Model:        dedalus.F("openai/gpt-image-1"),
        Background:   dedalus.F(dedalus.CreateImageRequestBackgroundTransparent),
        OutputFormat: dedalus.F(dedalus.CreateImageRequestOutputFormatPng),
    },
})

Download and Save Image

import (
    "io"
    "net/http"
)

response, err := client.Images.Generate(ctx, dedalus.ImageGenerateParams{
    CreateImageRequest: dedalus.CreateImageRequestParam{
        Prompt: dedalus.F("A beautiful sunset over the ocean"),
        Model:  dedalus.F("openai/dall-e-3"),
    },
})

if err != nil {
    log.Fatal(err)
}

// Download the image
resp, err := http.Get(response.Data[0].URL)
if err != nil {
    log.Fatal(err)
}
defer resp.Body.Close()

// Save to file
file, err := os.Create("generated_image.png")
if err != nil {
    log.Fatal(err)
}
defer file.Close()

_, err = io.Copy(file, resp.Body)
if err != nil {
    log.Fatal(err)
}

fmt.Println("Image saved successfully!")

Model Comparison

DALL-E 2

Fast and cost-effective
Good for simple images
Multiple images per request (n=1-10)
Lower resolution options available

DALL-E 3

Higher quality and more detailed
Better prompt adherence
Revised prompt feature
Single image per request (n=1)
HD quality option

GPT Image 1

Latest model with most features
Transparent backgrounds
Custom output formats and compression
Streaming support
Flexible quality levels

Best Practices

Prompt Writing: Be specific and descriptive in your prompts
Model Selection: Use DALL-E 2 for quick iterations, DALL-E 3 or GPT Image 1 for final high-quality images
Batch Generation: Use DALL-E 2 with n>1 for multiple variations
File Handling: Download and save images immediately as URLs expire after 60 minutes
Error Handling: Handle content policy violations gracefully

Overview

Chat

Embeddings

Audio

Images

Models

Overview

Method Signature

Request Parameters

Response Fields

Code Examples

Basic Image Generation

Generate Multiple Images

High Quality with Specific Style

Base64 Encoded Image

GPT Image 1 with Advanced Features

Transparent Background

Download and Save Image

Model Comparison

DALL-E 2

DALL-E 3

GPT Image 1

Best Practices

Build docs developers (and LLMs) love

Overview

Chat

Embeddings

Audio

Images

Models

​Overview

​Method Signature

​Request Parameters

​Response Fields

​Code Examples

​Basic Image Generation

​Generate Multiple Images

​High Quality with Specific Style

​Base64 Encoded Image

​GPT Image 1 with Advanced Features

​Transparent Background

​Download and Save Image

​Model Comparison

​DALL-E 2

​DALL-E 3

​GPT Image 1

​Best Practices

Build docs developers (and LLMs) love

Overview

Method Signature

Request Parameters

Response Fields

Code Examples

Basic Image Generation

Generate Multiple Images

High Quality with Specific Style

Base64 Encoded Image

GPT Image 1 with Advanced Features

Transparent Background

Download and Save Image

Model Comparison

DALL-E 2

DALL-E 3

GPT Image 1

Best Practices