Learn how to create stunning AI-generated images using the Weights API. This guide walks you through the complete process from creating an image generation job to retrieving your final results. You can create images with LoRas or with other images as a reference.

​
Step 1: Set Up Your Environment

First, install the Weights SDK and set up your authentication: 
npm install @weights-ai/sdk

import Weights from "@weights-ai/sdk";

const client = new Weights({
  bearerToken: "your-api-key-here",
});

​
Step 2: Create Your First Image

Start by creating a simple image generation job: 
const imageJob = await client.images.create({
  prompt: "A serene mountain landscape at sunset with golden clouds",
  dimensions: "LANDSCAPE",
  numImages: 1,
});

console.log(`Image job created with ID: ${imageJob.id}`);
console.log(`Status: ${imageJob.status}`);

​
Request Parameters

  • prompt (required): Your text description of the image you want to generate
  • dimensions (optional): Choose from SQUARE, PORTRAIT, or LANDSCAPE (default: SQUARE)
  • numImages (optional): Number of images to generate (1-4, default: 1)
  • loraIds (optional): Array of LoRA model IDs to use for the image generation

​
Step 3: Check Job Status

Image generation can be asynchronous. Poll the status to know when your image is ready: 
async function checkImageStatus(jobId:string) {
  const job = await client.images.retrieve(jobId);

  console.log(`Status: ${job.status}`);

  switch (job.status) {
    case "QUEUED":
      console.log("Your job is waiting in the queue...");
      break;
    case "PROCESSING":
      console.log("Your image is being generated...");
      break;
    case "SUCCEEDED":
      console.log("Image generated successfully!");
      console.log("Download URLs:", job.outputUrls);
      break;
    case "ERRORED":
      console.log("Generation failed. Please try again.");
      break;
  }

  return job;
}

// Check status every 3 seconds
const job = await checkImageStatus(imageJob.id);

​
Job Statuses

  • QUEUED: Job is waiting in the processing queue
  • PENDING_WORKER: Job is assigned to a worker
  • PROCESSING: Image is being generated
  • SUCCEEDED: Generation completed successfully
  • ERRORED: Generation failed
  • CANCELED: Job was canceled

​
Step 4: Retrieve Your Generated Images

Once the status is SUCCEEDED, your images are ready: 
if (job.status === "SUCCEEDED") {
  console.log("Your images are ready!");

  job.outputUrls.forEach((url, index) => {
    console.log(`Image ${index + 1}: ${url}`);
  });
}

​
Step 5: List All Your Images

View all your previous image generation jobs: 
const allImages = await client.images.list({
  limit: 25,
});

console.log(`You have ${allImages.jobs.length} image jobs:`);

allImages.jobs.forEach((job) => {
  console.log(`- ${job.id}: ${job.prompt} (${job.status})`);
  if (job.status === "SUCCEEDED" && job.outputUrls.length > 0) {
    console.log(`  Download: ${job.outputUrls[0]}`);
  }
});

​
Advanced Features

​
Using Custom LoRA Models

You can use trained LoRA models for specific styles: 
const customImage = await client.images.create({
  prompt: "A portrait of a woman in anime style",
  loraIds: [loraModel.id],
  dimensions: "PORTRAIT",
  numImages: 1,
});

​
Batch Generation

Generate multiple images at once: 
const batchJob = await client.images.create({
  prompt: "A futuristic city skyline at night",
  numImages: 4,
  dimensions: "LANDSCAPE",
});

​
Best Practices

​
Writing Effective Prompts

  • Be specific: “A majestic dragon with emerald scales flying over a medieval castle” vs “a dragon”
  • Include style details: “Oil painting style”, “photorealistic”, “anime style”
  • Specify lighting: “Golden hour lighting”, “dramatic shadows”
  • Mention composition: “Close-up portrait”, “wide landscape shot”

​
Error Handling

try {
  const imageJob = await client.images.create({
    prompt: "Your prompt here",
  });
} catch (error) {
  if (error.code === "BAD_REQUEST") {
    console.log("Invalid prompt or parameters");
  } else if (error.code === "UNAUTHORIZED") {
    console.log("Please check your API key");
  } else {
    console.log("An error occurred:", error.message);
  }
}

​
Complete Example

Here’s a complete example that creates an image and waits for completion: 
import Weights from "@weights-ai/sdk";

const client = new Weights({
  bearerToken: "your-api-key-here",
});

async function generateImage(prompt: string) {
  try {
    // Create the image generation job
    console.log("Creating image generation job...");
    const job = await client.images.create({
      prompt: prompt,
      dimensions: "LANDSCAPE",
      numImages: 1,
    });

    console.log(`Job created with ID: ${job.id}`);

    // Poll for completion
    let completed = false;
    while (!completed) {
      await new Promise((resolve) => setTimeout(resolve, 3_000)); // Wait 3 seconds

      const status = await client.images.retrieve(job.id);
      console.log(`Status: ${status.status}`);

      if (status.status === "SUCCEEDED") {
        console.log("Image generated successfully!");
        console.log("Download URL:", status.outputUrls[0]);
        completed = true;
      } else if (status.status === "ERRORED") {
        console.log("Generation failed");
        completed = true;
      }
    }
  } catch (error) {
    console.error("Error:", error.message);
  }
}

// Generate an image
generateImage("A magical forest with glowing mushrooms and fairy lights");

​
Next Steps

​
API Reference

For detailed API documentation, see the API Reference section.