Image Generation with DALL·E and Stable Diffusion Using APIs

AI-generated images are no longer science fiction. With models like DALL·E and Stable Diffusion, you can turn plain text into stunning visuals using simple API calls. Whether you’re building a web app, mobile project, or creative tool, these models give you immense flexibility.

In this guide, I’ll walk you through:

• How to use DALL·E and Stable Diffusion via APIs (OpenAI, Replicate, Stability.ai)

• Writing prompts that give you clean, accurate outputs

• Displaying or handling the generated images in your frontend project

Prerequisites

Before you begin, make sure you have:

• A developer account on OpenAI, Replicate, or Stability.ai

• API keys for your preferred provider

• Basic knowledge of JavaScript or Python

• A working frontend (React/Angular/Vue) or backend (Node/Go/Python) project

Step 1: Choosing Between DALL·E and Stable Diffusion

Model	Provider	Notes
DALL·E 3	OpenAI	High-quality, better for complex scenes
Stable Diffusion 1.5 / 2.1 / XL	Stability.ai, Replicate	Faster, customizable, open-source

For quick experiments or high-quality control, Stable Diffusion via Replicate is often a good choice. For commercial projects, DALL·E 3 with OpenAI’s API gives cleaner results.

 

Step 2: Designing Better Prompts

Prompt quality affects the output significantly. Here are some examples:

✅ Good Prompts:

• “A cyberpunk city at night, neon lights, 4K, cinematic lighting”

• “A hand-drawn sketch of a flying car, white background, top view”

🚫 Avoid Vague Prompts:

• “A nice image” ❌

• “Something cool” ❌

Use descriptive keywords like style, lighting, composition, and camera angle.

Step 3: Generating Images via API in Go

Using OpenAI’s DALL·E API (Go Example)

Install the OpenAI Go SDK (if not already done):

go get github.com/sashabaranov/go-openai

package main

import (
	"context"
	"fmt"
	"log"
	"os"

	openai "github.com/sashabaranov/go-openai"
)

func main() {
	client := openai.NewClient(os.Getenv("OPENAI_API_KEY"))

	req := openai.ImageRequest{
		Prompt:         "A medieval castle on a cliff during sunset",
		N:              1,
		Size:           "1024x1024",
		ResponseFormat: openai.CreateImageResponseFormatURL,
	}

	resp, err := client.CreateImage(context.Background(), req)
	if err != nil {
		log.Fatalf("Failed to generate image: %v", err)
	}

	fmt.Println("Generated image URL:", resp.Data[0].URL)
}

Make sure you set the OPENAI_API_KEY environment variable.

Using Replicate API (Stable Diffusion via Go HTTP client)

package main

import (
	"bytes"
	"encoding/json"
	"fmt"
	"io"
	"log"
	"net/http"
	"os"
)

func main() {
	apiKey := os.Getenv("REPLICATE_API_TOKEN")
	url := "https://api.replicate.com/v1/predictions"

	payload := map[string]interface{}{
		"version": "a9758cb3...your_model_version_id...",
		"input": map[string]string{
			"prompt": "A futuristic train station, cyberpunk style",
		},
	}

	data, _ := json.Marshal(payload)
	req, _ := http.NewRequest("POST", url, bytes.NewBuffer(data))
	req.Header.Set("Authorization", "Token "+apiKey)
	req.Header.Set("Content-Type", "application/json")

	client := &http.Client{}
	res, err := client.Do(req)
	if err != nil {
		log.Fatal("Request error:", err)
	}
	defer res.Body.Close()

	body, _ := io.ReadAll(res.Body)
	fmt.Println("Response:", string(body))
}

This will return a JSON response, including an image URL once processing is complete. You can poll for the result using the returned GET endpoint.